uva-dare (digital academic repository) microcirculatory ... · 16. derriford hospital, plymouth...

UvA-DARE is a service provided by the library of the University of Amsterdam (http://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

Microcirculatory dysfunction in critically ill patients: prevalence and significance from abedside perspective

Vellinga, N.A.R.

Link to publication

Citation for published version (APA):Vellinga, N. A. R. (2014). Microcirculatory dysfunction in critically ill patients: prevalence and significance from abedside perspective.

General rightsIt 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),other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, statingyour reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Askthe Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam,The Netherlands. You will be contacted as soon as possible.

Download date: 09 Jul 2020

https://dare.uva.nl/personal/pure/en/publications/microcirculatory-dysfunction-in-critically-ill-patients-prevalence-and-significance-from-a-bedside-perspective(66a0a8b6-39c5-4f64-9b68-21bb4aa7f808).html

CHAPTER 4International study on Microcirculatory Shock Occurrence in Acutely ill Patients (microSOAP) Accepted for publication in Critical Care Medicine

N.A.R. Vellinga1,2; E.C. Boerma2; M. Koopmans2; A. Donati3; A. Dubin4; N. I. Shapiro5; R.M. Pearse, 6; F.R. Machado7; M. Fries8; T. Akarsu-Ayazoglu9; A. Pranskunas10; S. Hollenberg11; G. Balestra12; M. van Iterson13; P. H. J. van der Voort14; F. Sadaka15; G. Minto16; U. Aypar17; F. J. Hurtado18; G. Martinelli19; D. Payen20; F. van Haren21; A. Holley22; R. Pattnaik23; H. Gomez24; R.L. Mehta25; A. H. Rodriguez26; C. Ruiz27; H.S. Canales28; J. Duranteau29; P. E. Spronk30; S. Jhanji31; S.M.A. Hubble32; M. Chierego33; C. Jung34; D .Martin35; C. Sorbara36; J. G.P. Tijssen37; J. Bakker1; C. Ince, PhD1

For the microSOAP study group (participating centers: see next page)

| Chapter 460

AFFILIATIONS:

1. Erasmus MC University Medical Center, Dept. of Intensive Care Adults, Rotterdam, the Netherlands

2. Medical Center Leeuwarden, Dept. of Intensive Care, Leeuwarden, the Netherlands

3. Università Politecnica delle Marche, Dept. of Biomedical Science and Public Health, Ancona, Italy

4. Sanatorio Otamendi y Miroli, Servicio de Terapia Intensiva, Azcuénaga 870, Buenos Aires, Argentina

5. Beth Isreal Deaconess Medical Center, Department of Emergency Medicine and Center for Vascular Biology Research, Boston, MA, United States of America

6. Barts and The London School of Medicine and Dentistry, London, United Kingdom

7. Dor e Terapia Intensiva, Universidade Federal de São Paolo, São Paolo, Brasil

8. Klinik für Anesthesiologie, Universitätsklinikum der RWTH Aachen, Aachen, Germany

9. K. Koșuyolu High Specialty Education and Research Hospital, Kartal Koșuyolu, Koșuyolu University, Istanbul, Turkey

10. Intensive Care Department, Lithuanian university of health sciences, Kaunas, Lithuania

11. Section of Cardiology, Cooper University Hospital, Camden, United States of America

12. Medical Intensive Care Unit, University Hospital Basel, Switzerland

13. Department of Anesthesiology, Intensive Care and Pain management, St. Antonius Hospital, Nieuwegein, the Netherlands

14. Department of Intensive Care, Onze Lieve Vrouwe Gasthuis, Amsterdam, the Netherlands

15. Critical Care Medicine/Neurocritical Care, Mercy Hospital St. Louis, St. Louis University Hospital, St. Louis, MO, United States of America

16. Derriford Hospital, Plymouth University Pensinsula School of Medicine, Plymouth, United Kingdom

17. Intensive Care Unit, Hacettepe University, Ankara, Turkey

18. Intensive Care Unit Hospital Español-ASSE, School of Medicine, UDELAR, Montevideo, Uruguay

19. Intensive Care Unit, New Cross Hospital, Wolverhampton, United Kingdom

20. Department of Anesthesiology, Critical Care et SMUR, Hôpital Lariboisière AP-HP/Université Paris 7 Diderot, Paris, France

21. Intensive Care Unit, Canberra Hospital, Canberra, Australia

22. Department of Intensive Care Medicine Royal Brisbane & Women’s Hospital, Brisbane, Australia

23. Intensive Care Unit, Ispat Hospital, Rourkela, Orissa, India

24. Critical Care Medicine, University of Pittsburgh, Pittsburgh, PA, United States of America

25. School of Medicine, University of California, San Diego, CA, United States of America

26. Critical Care Department, Joan XXIII University Hospital, Tarragona, Spain

4

61MicroSOAP: main results |

27. Departamento de Medicina Intensiva, Escuela de Medicina, Facultad de Medicina, Universidad Católica de Chile, Santiago, Chile

28. Intensive Care Unit, Hospital San Martín, La Plata, Argentina

29. Departement d’Anesthesie-Reanimation, Hôpitaux Universitaires Paris-Sud, Université Paris-Sud, Hôpital de Bicêtre AP-HP, Le Kremlin-Bicêtre, Paris, France

30. Intensive Care Unit, Gelre Ziekenhuizen, Apeldoorn, the Netherlands

31. Intensive Care Unit, The Royal Marsden Hospital, London, United Kingdom

32. Intensive Care Unit, Royal Devon and Exeter Hospital, Exeter, United Kingdom

33. Intensive Care Unit, Santa Maria degli Angeli Hospital, Pordenone, Italy

34. Universitätsklinikum Jena, Friedrich-Schiller-University, Department of Internal Medicine I, Jena, Germany

35. Intensive Care Unit, Royal Free Hospital, London, United Kingdom

36. Dipartimento di Anestesia, Rianimazione e Terapia Intensiva, Azienda ULSS 9 Veneto, Treviso, Italy

37. Academic Medical Center, Department of Cardiology, Amsterdam, the Netherlands

PARTICIPATING CENTERS AND MEMBERS OF THE MICROSOAP STUDY GROUP:

1. ICU, Medical Center Leeuwarden, the Netherlands (E.C. Boerma, MD, PhD, M. Koopmans, RN; N.A.R. Vellinga, MD)

2. ICU, St. Antonius Hospital, Nieuwegein, the Netherlands (M. van Iterson, MD, PhD)

3. ICU, Onze Lieve Vrouwe Gasthuis, Amsterdam, the Netherlands (P.H.J. van der Voort, MD, PhD)

4. ICU, Erasmus Medical Center, Rotterdam, the Netherlands (J. Bakker, MD, PhD; J. van Bommel, MD, PhD; C. Ince PhD)

5. ICU, Gelre Ziekenhuizen, Apeldoorn, the Netherlands (P.E. Spronk, MD, PhD, FCCP)

6. Departamento de Medicina Intensiva, Escuela de Medicina, Facultad de Medicina, Pontificia Universidad Católica de Chile, Santiago, Chile (C. Ruiz, MD; G. Hernandez, MD, PhD)

7. Departamento de Anestesiologia, Dor e terapia Intensiva, Hospital Sao Paulo, Universidade Federal de São Paulo Sao Paulo, Brasil (F.R. Machado, MD, PhD; A.T. Bafi, MD)

8. Servicio de Terapia Intensiva, Sanatorio Otamendi y Miroli, Buenos Aires, Argentina (A. Dubin, MD, PhD; V.S. Kanoore Edul, MD)

9. ICU, Hospital San Martín, La Plata, Argentina (H.S. Canales, MD)

10. ICU, Hospital Español ‘Juan J Crotoggini’, Montevideo, Uruguay (F.J. Hurtado, MD; G. Lacuesta, MD; M. Baz, MD)

11. ICU, Cooper University Hospital, Cooper Medical School of Rowan University, Camden, USA (S.M. Hollenberg MD, FACC, FCCM, FAHA, FCCP; U. Patel, MD)

| Chapter 462

12. ICU, Beth Israel Deaconess Medical Center/Harvard Medical School, Boston USA (N.I. Shapiro, MD, MPH)

13. Critical Care Medicine, University of Pittsburgh, Pittsburgh, PA, USA (H. Gomez, MD; P. Simon, MD; M. Pinsky, MD, CM, Dr hc, MCCM, FCCP)

14. Critical Care Medicine/Neurocritical Care, Mercy Hospital St Louis/ St. Louis University Hospital, St Louis, Missouri, USA (F.G. Sadaka, MD; K. Krause, RN)

15. ICU, University of California, San Diego, USA ( R. Mehta, MD, PhD)

16. Universitätsklinikum Jena, Friedrich-Schiller-University, Department of Internal Medicine I, Jena Germany (C. Jung, MD)

17. Department of Surgical Intensive Care, University Hospital Aachen, Germany (M. Fries, MD, PhD)

18. Adult Critical Care Unit, Royal London Hospital, London, UK (R.M. Pearse, MBBS, FRCA, FFICM; A. Smith, RGN)

19. ICU, Royal Free Hospital, London, UK (D.S. Martin, MD, PhD; P. Meale, RGN)

20. ICU, The Royal Marsden Hospital, Chelsea, London, UK (S. Jhanji, MD, PhD)

21. ICU, Derriford Hospital, Plymouth; Plymouth University Peninsula School of Medicine, Plymouth, UK (G. Minto, MD, FRCA; C. Lai , C. Ferguson, H. McMillan, T. Quintrell, M. Sair)

22. ICU, New Cross Hospital, Wolverhampton, UK (G. Martinelli, MD; M. Lombrano, MD)

23. ICU, Royal Devon and Exeter Hospital, Exeter, UK (S.M.A. Hubble, MD; C. Thorn, PhD)

24. Critical Care Department, Joan XXIII University Hospital, Tarragona, Spain (A.H. Rodriguez, MD, PhD; I. Martin-Loeches, MD, PhD (current affiliation: Critical Care Centre, Corporacio Sanitaria I Universitaria Parc Tauli – Hospital de Sabadell, Barcelona, Spain))

25. Department of Intensive Care Medicine, Waikato Hospital, Hamilton, New Zealand (F.M.P. van Haren, MD, PhD (current affiliation: Intensive Care Unit, Canberra Hospital, Canberra, Australia)

26. Intensive Care Department, Lithuanian University of Health Sciences, Kaunas, Lithuania (A. Pranskunas, MD, PhD; V. Pilvinis, MD, PhD)

27. Clinica di Anestesia e Rianimazione, Azienda Ospedaliera-Universitaria Ospedali Riuniti, Ancona, Italy (A. Donati, MD)

28. Dipartimento di Anestesia, Rianimazione e Terapia Intensiva, Azienda ULSS 9 Veneto, Treviso, Italy (C. Sorbara, MD; A. Forti, MD, A. Comin, PhD)

29. ICU, Santa Maria degli Angeli Hospital, Pordenone, Italy (M.L. Chierego, MD; T. Pellis, MD)

30. ICU, The University of Queensland and Royal Brisbane and Women’s Hospital, Brisbane, Australia (A. Holley, MD, FACEM, FCICM; J. Paratz, MD, PhD)

31. Departement d’Anesthesie-Reanimation, Hôpitaux Universitaires Paris-Sud, Université Paris-Sud, Hôpital de Bicêtre AP-HP, Le Kremlin-Bicêtre, France, Paris, France (J. Duranteau, MD, PhD; A. Harrois, MD)

32. Department of Anesthesiology, Critical Care et SMUR, Hôpital Lariboisière AP-HP/Université Paris 7 Diderot, Paris, France (D. Payen, MD, PhD; M. Legrand, MD, PhD)

4


33. Medical ICU, University Hospital Basel, Switzerland (G.M. Balestra, MD; E. Bucher, MD)

34. Ispat Hospital, Rourkela, Orissa, India (R. Pattnaik) & Faculty of Tropical Medicine, Mahidol University, Bangkok, Thailand (A.M. Dondorp, MD, PhD; M.T. Herdman, MD, PhD)

35. ICU, Hacettepe University, Ankara, Turkey (U. Aypar MD, B. Ayhan, MD)

36. ICU, Kosuyolu University, Istanbul, Turkey (T. Ayazoglu-Akarsu, MD)

The work was performed in the intensive care units of the hospitals named above in the microSOAP study group affiliations and was coordinated from the Intensive Care Unit of Medical Center Leeuwarden, Leeuwarden, the Netherlands.

Financial support and potential conflicts of interest:Supported, in part, by an unrestricted grant from the local hospital fund, Medical Center Leeuwarden, Leeuwarden, the Netherlands. The funder had no role in the study design and data acquisition, analysis, interpretation and review, or approval of the manuscript. No financial compensation was received by participating centers or persons who made additional contributions.

Dr. Vellinga’s and Dr. Boerma’s institution received grant support from a local hospital fund (Medical Center Leeuwarden, unrestricted grant). Dr. Shapiro served as board member for Cumberland DSMB. His institution received grant support from Biosite, Cheetah Medical, Rapid Pathogen Screening, and Thermo-Fisher. Dr. Pearse consulted for Massimo and Edwards Lifesciences and lectured for Nestle Health Sciences. His institution received grant support from Nestle Health Sciences, LiDCO, and Cephalon (some are equipment loans not funding). Dr. Payen consulted for Vygon Italy. Dr. Mehta consulted for Abbvie CSL Behring, AM Pharma, Grifols, Ardea, and GlaxoSmithKline; provided expert testimony for Nell DyMott; lectured for Abbvie; and has stock options with Astute. His institution received grant support from Spectral, Allocure, and Eli Lilly. Dr. Rodriguez served as board member for MSD; consulted and lectured for Pfizer, Astellas, Novartis, and Brahms; and received support for travel from Pfizer, MSD, Astellas, Novartis, and Brahms. Dr. Canales has disclosed government work. Dr. Duranteau’s institution lectured for the LFB company. Dr. Hubble is employed by Royal Devon and Exeter Hospital NHS trust (intensive care consultant). Dr. Sorbara has disclosed government work. Dr. Ince is the inventor of Sidestream Dark Field technology, which is commercialized by MicroVision Medical. He has been a consultant for this company in the past, but he has broken all contact with this company for more than 4 years now, and he has no competing interests other than his commitment to promote the importance of the microcirculation in the care of critically ill patients. He has stock options in MicroVision Medical Trust. The remaining authors have disclosed that they do not have any potential conflicts of interest.

| Chapter 464

Author’s contributions: Drs. Vellinga, Boerma, Koopmans, Donati, Dubin, Shapiro, Pearse, Bakker and Ince conceived and designed the study. Data acquisition was peformed by: Vellinga, Boerma, Koopmans, Donati, Dubin, Shapiro, Pearse, Machado, Fries, Akarsu-Ayazoglu, Pranskunas, Hollenberg, Balestra, Van Iterson, Van der Voort, Sadaka, Minto, Aypar, Hurtado, Martinelli, Payen, Van Haren, Holley, Pattnaik, Gomez, Mehta, Rodriguez, Ruiz, Canales, Duranteau, Spronk, Jhanji, Hubble, Chierego, Jung, Martin, Sorbara, Bakker, Ince. Vellinga, Boerma, Koopmans, Tijssen and Ince were responsible for data analysis. Tijssen provided statistical expertise. Vellinga, Boerma, Koopmans, Donati, Dubin, Shapiro, Pearse, Tijssen, Bakker and Ince interpreted the data. Vellinga, Boerma and Ince wrote the manuscript draft. All authors revised the manuscript for important intellectual content and approved the article.

Keywords: microcirculation, Sidestream Dark Field imaging, in-vivo microscopy, tachycardia

4


ABSTRACT

Objectives Microcirculatory alterations are associated with adverse outcome in subsets of critically ill patients. The prevalence and significance of microcirculatory alterations in the general intensive care unit (ICU) population are unknown. We studied the prevalence of microcirculatory alterations in a heterogeneous ICU population and its predictive value in an integrative model of macro- and microcirculatory variables.

Design Multicenter observational point prevalence study.

Setting The Microcirculatory Shock Occurrence in Acutely ill Patients (microSOAP) study was conducted in 36 ICUs worldwide.

Patients A heterogeneous ICU population consisting of 501 patients.

Interventions None.

Measurements and main resultsDemographic, hemodynamic and laboratory data were collected in all ICU patients ≥18 years. Sublingual Sidestream Dark Field imaging was performed to determine the prevalence of an abnormal capillary microvascular flow index (MFI < 2.6) and its additional value in predicting hospital mortality.In 501 patients with a median APACHE II score of 15 [10-21], a SOFA score of 5 [2-8] and a hospital mortality of 28.4%, 17% exhibited an abnormal capillary MFI. Tachycardia (heart rate >90 bpm) (odds ratio (OR) 2.71, 95%-confidence interval (CI) 1.67-4.39, P<0.001), mean arterial pressure (OR 0.979, 95%-CI 0.963-0.996, P=0.013), vasopressor use (OR 1.84, 95%-CI 1.11-3.07, P=0.019) and lactate level >1.5 mEq/L (OR 2.15, 95%-CI 1.28-3.62, P=0.004) were independent risk factors for hospital mortality, but not abnormal MFI. In reference to MFI, a significant interaction was observed with tachycardia. In tachycardic patients, the presence of an abnormal MFI was an independent, additive predictor for in-hospital mortality (OR 3.24, 95%-CI 1.30-8.06, P=0.011). This was not true for non-tachycardic patients, nor for the total group of patients.

Conclusions:In a heterogeneous ICU population, an abnormal MFI was present in 17% of patients. This was not associated with mortality. However, in patients with tachycardia, an abnormal MFI was independently associated with an increased risk of hospital death.

| Chapter 466

INTRODUCTION

The presence and significance of microcirculatory alterations in the early phase of critical illness, including sepsis and heart failure, has been widely explored [1-4]. Although various techniques can provide information on microvascular dysfunction, discrimination of capillary and venule perfusion appears to be of paramount importance and relies on direct in vivo microscopy methods, including Sidestream Dark Field (SDF) imaging [5-7]. Sublingual microcirculatory abnormalities identified by SDF are considered clinically relevant and are independently associated with an increased risk of morbidity and mortality [1,2,8-13]. Conventional hemodynamic monitoring appears to fall short in detecting this ‘microcirculatory shock’: a common finding is the absence of a clear association between the microcirculation and macrohemodynamic variables, such as cardiac output and blood pressure [8,9,11-18]. Therefore, the microcirculation has the potential to be an important additional target for monitoring both organ perfusion and treatment efficacy [3,19-21]. Although conventional goal-directed therapy is associated with improvement of capillary perfusion, persisting microcirculatory abnormalities, despite fulfillment of resuscitation endpoints, are related to adverse outcome [1,12,13]. Interventions intended to ameliorate microcirculatory dysfunction have shown varying results and lack a clear association with improved outcome [18,20,22-25]. To further understand the role of microcirculatory monitoring and microcirculation-directed interventions, knowledge of the prevalence of microcirculatory alterations in the general intensive care population is of utmost importance. To date, our knowledge is predominantly based on single-center studies in high mortality subgroups in the early phases of critical illness. This implies that data on the prevalence of microcirculatory alterations in the general, heterogeneous intensive care setting are not currently available. Observational studies in multicenter settings, such as the Sepsis Occurrence in Acutely ill Patients (SOAP) and European Prevalence of Infection in intensive Care (EPIC) trials, are valuable tools and have contributed greatly to our knowledge of the prevalence and incidence of diseases [26,27]. We applied a similar study design, focusing on current ICU patient characteristics and hemodynamic monitoring, in a worldwide multicenter setting. Furthermore, we evaluated the prevalence and prognostic value of microcirculatory alterations in our heterogeneous ICU population. In this paper, we present our main findings.

MATERIALS AND METHODS

Patient inclusionThe microSOAP study (NCT01179243) trial was scheduled for September 5-9, 2011 [28]. ICU patients ≥ 18 years of age, regardless of their underlying disease, were eligible for inclusion. All centers obtained medical ethics approval (or a waiver, if applicable). Written informed consent for included subjects was obtained in accordance with local applicable laws. The exclusion criteria were a lack of informed consent and patient-related factors that substantially interfered with SDF imaging, such as recent maxillofacial surgery or mucosal bleeding or injury. Funding consisted of an unrestricted grant from a local hospital fund.

4


Data collectionThe data on patient characteristics, hemodynamics, laboratory values and treatment were collected together with simultaneous SDF imaging of the sublingual microcirculation. Being a point prevalence study, data were collected on the same day for all patients in a given ICU or ICU subunit.

SDF imagingThe noninvasive SDF technique consists of a handheld camera, emitting stroboscopic green light with a wavelength within the absorption spectrum of hemoglobin [5]. When placed on mucosal surfaces, the stroboscopic light is absorbed by hemoglobin, thereby visualizing blood vessels by depicting erythrocytes as black dots [29-31]. Images were obtained and analyzed in agreement with internationally accepted consensus [29,30].

Data analysis

SDF analysisSDF clips were blindly analyzed offline in a random order by a preselected group of well-trained SDF researchers. Aiming for consensus, images were excluded in cases of pressure artifacts, instability or inadequate focus that substantially interfered with the analysis. The coefficient of variation for the analysis was calculated based on 10 randomly selected SDF images.

Computer-assisted analysis (AVA 3.0 software, MicroVision Medical, Amsterdam, the Netherlands) was performed in line with international consensus. The semi-quantitative microvascular flow index (MFI), ranging from 0 (no flow) to 3 (continuous flow), and percentage of perfused vessels (PPV) provide information on convexity. MFI is scored as the predominant type of flow in every image quadrant for every image. For each patient, the average MFI is calculated [12,29,30,32]. Total and perfused vessel density (TVD, PVD), both in mm/mm2, provide information on diffusion. The image analysis is described in detail elsewhere [29,30]. Being the minimum reported value for the lower bound of the 95%-confidence interval (CI) of MFI in healthy volunteers, a small vessel (<20 µm) MFI<2.6 was considered as abnormal [4,12,33,34]. Microcirculatory variables pertain to small vessels, unless indicated otherwise.

Statistical analysisPatient data were described using descriptive statistics. Student’s t-test, the Mann-Whitney U test or Fisher’s exact test were used to test for differences between variables. Backwards stepwise multivariable logistic regression analysis was applied to identify predictors of hospital mortality. For multivariable models, an area under the curve (AUC) was calculated. Logistic regression analysis was repeated in patients with tachycardia, a post hoc defined subgroup based on a significant interaction between tachycardia (heart rate > 90 beats per minute) and an abnormal MFI. To correct for unavailability of data, multiple imputation analysis was used (20 imputations). Lactate values and microcirculatory parameters were not imputated. Hosmer & Lemeshow goodness of fit test was used to

| Chapter 468

describe the fit of the model. Statistical analysis is described in detail in the Supplemental Content. The data were analyzed using SPSS 21.0 (IBM, New York, USA) and GraphPad Prism 5.04 (GraphPad Software, Inc., La Jolla, USA) and are presented as the median [interquartile range] or mean ± standard deviation, unless indicated otherwise. A P<0.05 was considered statistically significant.

RESULTS

All patients

Patient inclusionOut of 753 screened patients, 531 patients were included from 36 ICUs worldwide (figure 1). The majority of exclusions (68%) were not SDF related, with 57% attributable to the lack of informed consent. Twenty patients (3.8%) were excluded because of insufficient SDF image quality. 501 patients (81% of the eligible patients) were available for further analysis.

Figure 1. Overview of screened, included and excluded patients. Reasons for exclusion are divided into SDF related and SDF non-related (including absence of informed consent). ICU = Intensive Care Unit, SDF = Sidestream Dark Field Imaging.

4


General characteristicsBaseline characteristics of the study population are shown in table S1, Supplemental Content. Patients were 62 [51-73] years old with an APACHE II score of 15 [10-21] and a SOFA score of 5 [2-8]. The most common reasons for ICU admission were surgery (33%) and sepsis (17%).

Microcirculatory variablesAn abnormal MFI was observed in 86 patients (17%). The number of adequately resuscitated patients, as determined by the attending physician, did not differ between patients with and without an abnormal MFI (85% vs. 78%, P=0.15). In patients with an abnormal MFI, we observed a higher heterogeneity index (0.80 [0.40-1.32] vs. 0.00 [0.00-0.35], P<0.001) and a lower PPV (0.92 [0.87-0.95] vs. 0.98 [0.96-1.00], P<0.001) and PVD (17.21 ± 3.95 vs. 18.83 ± 3.88 mm/mm2); however, TVD did not differ (18.93 ± 3.98 vs. 19.32 ± 3.99 mm/mm2, P=0.41) (Supplemental Content). No differences in microcirculatory variables were observed between different admission diagnoses. The coefficient of variation for the SDF analysis varied from 0%±0% for large vessel MFI and 2%±2% for small vessel MFI to 7%±4% for the (perfused) De Backer score.

OutcomeHospital non-survivors (28.4%) displayed higher APACHE II and SOFA scores, lower hemoglobin, and a higher heart rate and arterial lactate level (table S2, Supplemental Content). Multivariable logistic regression identified higher APACHE II score, a stay in ICU > 24 hours before SDF, arterial lactate level > 1.5 mEq/L, tachycardia, lower MAP, renal replacement therapy, use of a vasopressor as well as being admitted to ICU because of sepsis, respiratory insufficiency or cardiac disease as independent predictors of hospital mortality (for odds ratios see table 1, AUC for this model 0.83, 95%-CI 0.79-0.87, P<0.001).

| Chapter 470

Table 1. Multivariable logistic regression for variables associated with hospital mortality

A. Hospital mortality OR (95% CI) PAPACHE II 1.05 (1.02-1.08) 0.001

Stay in ICU > 24 hours before SDF 2.81 (1.53-5.21) 0.001

Lactate level > 1.5 mEq/l* 2.15 (1.28-3.62) 0.004

Heart rate > 90 bpm 2.71 (1.67-4.39) <0.001

MAP (mmHg) 0.979 (0.963-0.996) 0.013

Use of any vasopressor 1.84 (1.11-3.07) 0.019

Renal replacement therapy 2.26 (1.07-4.80) 0.034

Reason for ICU admission** 0.022

Surgery (reference category) 1.00

Sepsis 2.07 (1.02-4.22) 0.045

Trauma/hemorrhage/other 1.58 (0.74-3.35) 0.239

Resp. insufficiency/cardiac disease 3.44 (1.73-6.87) <0.001

Neurological disorders 2.32 (0.96-5.58) 0.061

B. Hospital mortality for HR>90 OR (95% CI) PLactate level > 1.5 mEq/l* 2.84 (1.36-5.92) 0.005


Abnormal MFI 3.24 (1.30-8.06) 0.011




Sepsis 3.08 (1.27-7.45) 0.013


Resp. insufficiency/cardiac disease 3.54 (1.38-9.08) 0.008


A. Multivariable logistic regression for hospital mortality (all patients). Average Nagelkerke R2 = 0.37 (range: 0.36-0.41) (P<0.001), average Hosmer & Lemeshow Chi-square 5.303 (range 1.383-9.400), P=0.710 (range 0.310-0.994) . B. Multivariable logistic regression for hospital mortality for patients with heart rate > 90/min (n= 204). Nagelkerke R2 = 0.32 (P<0.001), Hosmer & Lemeshow Chi-square 5.576, P=0.59. P=0.022 for overall effect of admission diagnosis. Because APACHE II score was not included in this model, models in all imputations were equal. OR = odds ratio, 95% CI = 95% confidence interval, ICU = intensive care unit, SDF = Sidestream Dark Field imaging, bpm = beats per minute, MAP = mean arterial pressure, abnormal MFI = MFI<2.6 for vessels < 20 µm.*As compared to patients with a lactate <1.5 mEq/L or no lactate measurement available. ** As compared to patients with surgery as admission diagnosis

Integrating micro- and macrohemodynamic monitoring

The use of macrohemodynamic monitoring other than blood pressure and heart rate appeared to be very limited (Table S3, Supplemental content). The measurement of cardiac output and S(c)vO2 was restricted to 6.2 and 20% respectively. Furthermore, the percentage of patients with hypotension, defined as a MAP < 65 mmHg was as low as

4


8%. However, tachycardia, defined as a heart rate (HR) >90 bpm, was present in 204 (41%) patients. This threshold was confirmed for our database as the optimal cut-off value for hospital mortality with a sensitivity of 66% and a specificity of 62% (AUC 0.69 (0.63-0.74), P<0.001). Tachycardia was significantly less frequent in patients who had been admitted to the ICU less than 24 hours prior, compared to other patients (47% vs. 27%, P<0.001). Tachycardic patients had significantly higher APACHE II and SOFA scores and lower hemoglobin levels. No significant differences were observed in lactate levels (1.3 [0.9-2.1] vs. 1.2 [0.9-2.0] mEq/L, P=0.28). Vasopressor use was more frequent in patients with a HR>90 bpm (37% vs. 26%, P=0.007) (Table S4, Supplemental Content). No significant difference was observed between patients with or without tachycardia in terms of the number of subjects considered adequately resuscitated (78% vs. 81%, P=0.50). In contrast, these data for patients with and without hypotension (MAP<65 mmHg) were 48% and 82%, respectively (P<0.001).

OutcomeTachycardia was an independent predictor of hospital mortality (HR>90 bpm 41%, HR≤90 bpm 19%, P<0.001; OR 2.71 (1.67-4.39), P<0.001) (table 1B, figure 2). In patients with tachycardia, not only lactate levels > 1.5 mEq/L, but also an abnormal MFI was one of the independent, additional risk factors for in-hospital death (68% vs. 38%, P=0.002; OR 3.24 (1.30-8.06), P=0.011, table 1). AUC (95% CI) for this model was 0.79 (0.73-0.86), P<0.001). In contrast, an abnormal MFI did not have an additional predictive value in patients with a HR≤90 bpm (figure 2).

Figure 2. Hospital mortality for subgroups of patients with and without tachycardia (heart rate > 90 beats per minute). HR, heart rate. Bpm, beats per minute. Normal MFI, i.e., MFI ≥2.6 for vessels < 20 µm. Abnormal MFI, MFI <2.6 for vessels < 20 µm. P <0.05 is considered statistically significant.

| Chapter 472

DISCUSSION

By including over 500 patients with a variety of underlying diseases, the microSOAP trial is presently the largest prospective study investigating the prevalence and significance of microcirculatory alterations in a heterogeneous ICU population. Applying a predefined threshold, an abnormal MFI was observed in 17% of patients [4]. In the mixed ICU population, lactate levels and several macrohemodynamic variables, but not microcirculatory variables, were independent predictors of hospital mortality. After post hoc identification of a high-risk tachycardic subpopulation, abnormal microcirculatory blood flow was an additional independent risk factor for death.

In our study, the likelihood of microcirculatory abnormalities was lower than reported previously. This may in part be explained by patient selection: our patients were less severely ill than the patients in previously studied subgroups. Moreover, the majority of studies are restricted to the early phase of critical illness. In the present study, the smaller number of patients in subgroups such as sepsis did not allow in depth subgroup analysis. Furthermore, abnormal microcirculatory blood flow was not an independent risk factor in the overall population, whereas a significant difference in abnormal microcirculatory blood flow variables has been observed between survivors and non-survivors in various studies. However, previous smaller studies primarily focused on high mortality subgroups. Indeed, in a recent study in early normotensive sepsis, MFI was 3.00 [2.73-3.00] [35].

The observed association between macrohemodynamic variables, lactate levels and mortality confirms the present clinical paradigm [36-46]. Notably, a single measurement of blood pressure or heart rate, irrespective of disease state and timeframe, provided predictive value.

The prognostic significance of tachycardia is well-recognized, especially in cardiac disease, but also in different phases of critical illness [44,45,47-50]. In line with previous literature and our data, we used a cut-off value of 90 bpm for further analysis [45,49-51]. Using this cut-off value at ICU discharge resulted in similar differences in mortality in patients with multiple organ dysfunction syndrome as observed in our study population[45].

In contrast to previous literature, we found indications for an association between macro- and microcirculatory variables: an abnormal MFI was an independent predictor for hospital death in subjects with tachycardia. This was independent of inotrope use. Linking microcirculatory abnormalities with hypotension was impossible due to the low prevalence (8%) of patients with a MAP <65 mmHg. However, tachycardia was present in 41% of patients and was not confined to patients included within the first 24 hours of ICU admission. Furthermore, the attending physician considered resuscitation adequate at the moment of data acquisition, irrespective of the presence of tachycardia. These data are in agreement with the fulfillment of resuscitation goals in the existing guidelines, in which heart rate is not an endpoint. In addition, some patients appear to display a well-compensated microcirculatory blood flow under conditions of increased stress, including tachycardia, whereas others do not. Persisting microcirculatory shock has been related to adverse outcome, and accordingly, patients in whom microcirculatory perfusion increases

4


during the course of their disease may have an increased chance of a better outcome [3,52]. Therefore, the ability to preserve microcirculatory perfusion under conditions of stress appears to be key to a more favorable clinical course.

This study has several limitations. A capillary MFI< 2.6 was a priori defined as the threshold for an abnormal MFI. MFI was chosen because of the possibility of bedside evaluation of this parameter, in contrast to the mandatory offline analysis for other parameters [53]. This could maximize the clinical applicability of the findings. The threshold value was based on previous studies describing the range of MFI in healthy volunteers [4,12,33,34]. In order to minimize false positive findings, the minimum reported lower bound of the 95% CI interval in healthy volunteers was used as threshold value [12]. A capillary MFI < 2.6 has been shown to be the optimal cut-off value for the response to fluid administration [19]. Although SDF enables direct visualization of capillaries, several other techniques such as near infrared spectroscopy and laser Doppler flowmetry are also useful in providing information on the microcirculation [6].

Due to the design of the study, data on the incidence of microcirculatory flow abnormalities cannot be estimated. Presumably, our data underestimate the true incidence of microcirculatory dysfunction, as it has been observed that these alterations attenuate over time [1,22,54]. Furthermore, no information on the relevance of changes of both macro- and microcirculatory variables over time can be provided. Because of a significant interaction between tachycardia and an abnormal MFI, analysis in the subgroup of tachycardic patients was appropriate, nevertheless being a post hoc analysis. Despite being the largest prospective study in this field this far, lower numbers of patients per subgroups may have masked clinically relevant differences.

Approximately one third of all screened patients were not included in the study, predominantly due to a lack of informed consent. However, the vast majority of reasons for non-inclusion were not related to sublingual in vivo microscopy, and only 3.8% of the included patients were excluded because of inadequate SDF image quality. In agreement with previous literature, the coefficient of variation for our analysis was good [9,12,30]. Moreover, by aiming for consensus between SDF researchers, we aimed to keep differences in analysis to a minimum. It must be mentioned, however, that although improved technology is forthcoming, the current need for detailed offline analysis is a severe impairment of the practical bedside applicability of this technique [55]. Finally, the limited macrohemodynamic monitoring did not allow for an extensive evaluation of a possible relationship between S(c)vO2 or cardiac output and microcirculatory variables. Our data reflect daily clinical practice in critical care and seem to be in contrast with an overwhelming interest for more advanced hemodynamic monitoring in the current literature .

| Chapter 474

CONCLUSIONS

The present study provides an estimation of the prevalence of microcirculatory abnormalities in a heterogeneous ICU population and may serve as a basis for future studies. In this general ICU population, an abnormal MFI is not associated with mortality, whereas the presence of an abnormal MFI independently predicts an increased risk of dying in patients already at risk for adverse outcome due to tachycardia . Our data bridge the gap between micro- and macrocirculatory dysfunction, suggesting that microcirculatory monitoring could be a potentially clinically important extension of conventional hemodynamic monitoring. Future research could seek to unravel the underlying mechanisms of microcirculatory shock and potential therapeutic options.

ACKNOWLEDGEMENTS

We thank prof. J.L. Vincent, Dept. of Intensive Care, Erasme University Hospital, Brussels, Belgium, for allowing us to use the SOAP acronym in our study; F. Messie, Osix, Huizen, the Netherlands, for software engineering; A. Carsetti, MD, R. Domizi, MD and C. Scorcella, MD, Università Politecnica delle Marche, Dept. of Biomedical Science and Public Health, Ancona, Italy; G. Veenstra, MD, and B. Scheenstra, MD, ICU Medical Center Leeuwarden, the Netherlands, for their contributions to SDF analysis. D. M. J. Milstein, PhD, and K. Yuruk, MD, Dept. of Translational Physiology, Academic Medical Center, Amsterdam, the Netherlands for their help with SDF imaging. P. Ormskerk, RN, and D. van Duijn, RN, Intensive Care Adults, Rotterdam, the Netherlands, for their help with study logistics. D.S. Martin, MD, PhD, P. Meale, RGN, University College London, ICU, Royal Free Hospital, London, United Kingdom; and A. Vivian-Smith, RGN, ICU, Royal London Hospital, London, United Kingdom, for their assistance with UK ethics approval.

4


REFERENCES

1. De Backer D, Donadello K, Sakr Y et al.: Microcirculatory Alterations in Patients With Severe Sepsis: Impact of Time of Assessment and Relationship With Outcome. Crit Care Med 2013, 41(3):791-799

2. den Uil CA, Lagrand WK, van der Ent M et al.: Impaired microcirculation predicts poor outcome of patients with acute myocardial infarction complicated by cardiogenic shock. Eur Heart J 2010, 31(24):3032-3039.

3. Trzeciak S, McCoy JV, Phillip Dellinger R et al.: Early increases in microcirculatory perfusion during protocol-directed resuscitation are associated with reduced multi-organ failure at 24 h in patients with sepsis. Intensive Care Med 2008, 34(12):2210-2217.

4. Edul VSK, Enrico C, Laviolle B et al.: Quantitative assessment of the microcirculation in healthy volunteers and in patients with septic shock. Crit Care Med 2012, 40(5):1443-1448.

5. Goedhart PT, Khalilzada M, Bezemer R et al.: Sidestream Dark Field (SDF) imaging: a novel stroboscopic LED ring-based imaging modality for clinical assessment of the microcirculation. Optics express 2007, 15(23):15101-15114.

6. De Backer D, Ospina-Tascon G, Salgado D et al.: Monitoring the microcirculation in the critically ill patient: current methods and future approaches. Intensive Care Med 2010, 36(11):1813-1825.

7. Groner W, Winkelman JW, Harris AG et al.: Orthogonal polarization spectral imaging: a new method for study of the microcirculation. Nat Med 1999, 5(10):1209-1212.

8. Jhanji S, Lee C, Watson D et al.: Microvascular flow and tissue oxygenation after major abdominal surgery: association with post-operative complications. Intensive Care Med 2009, 35(4):671-677.

9. De Backer D, Creteur J, Preiser JC et al.: Microvascular blood flow is altered in patients with sepsis. Am J Respir Crit Care Med 2002, 166(1):98-104.

10. Top APC, Ince C, de Meij N et al.: Persistent low microcirculatory vessel density in nonsurvivors of sepsis in pediatric intensive care. Crit Care Med 2011, 39(1):8-13.

11. De Backer D, Creteur J, Dubois MJ et al.: Microvascular alterations in patients with acute severe heart failure and cardiogenic shock. Am Heart J 2004, 147(1):91-99.

12. Trzeciak S, Dellinger RP, Parrillo JE et al.: Early microcirculatory perfusion derangements in patients with severe sepsis and septic shock: relationship to hemodynamics, oxygen transport, and survival. Ann Emerg Med 2007, 49(1):88-98, 98.e1.

13. van Genderen ME, Lima A, Akkerhuis M et al.: Persistent peripheral and microcirculatory perfusion alterations after out-of-hospital cardiac arrest are associated with poor survival*. Crit Care Med 2012, 40(8):2287-2294.

14. De Backer D, Ortiz JA, Salgado D: Coupling microcirculation to systemic hemodynamics. Curr Opin Crit Care 2010, 16(3):250-254.

15. De Backer D, Dubois MJ, Schmartz D et al.: Microcirculatory alterations in cardiac surgery: effects of cardiopulmonary bypass and anesthesia. Ann Thorac Surg 2009, 88(5):1396-1403.

| Chapter 476

16. Hernandez G, Bruhn A, Castro R et al.: Persistent Sepsis-Induced Hypotension without Hyperlactatemia: A Distinct Clinical and Physiological Profile within the Spectrum of Septic Shock. Critical care research and practice 2012, 2012:536852.

17. Boerma EC, Kuiper MA, Kingma WP et al.: Disparity between skin perfusion and sublingual microcirculatory alterations in severe sepsis and septic shock: a prospective observational study. Intensive Care Med 2008, 34(7):1294-1298.

18. Morelli A, Donati A, Ertmer C et al.: Levosimendan for resuscitating the microcirculation in patients with septic shock: a randomized controlled study. Crit Care 2010, 14(6):R232.

19. Pranskunas A, Koopmans M, Koetsier PM et al.: Microcirculatory blood flow as a tool to select ICU patients eligible for fluid therapy. Intensive Care Med 2013, 39(4):612-619.

20. Jhanji S, Stirling S, Patel N et al.: The effect of increasing doses of norepinephrine on tissue oxygenation and microvascular flow in patients with septic shock. Crit Care Med 2009, 37(6):1961-1966.

21. Dubin A, Pozo MO, Casabella CA et al.: Increasing arterial blood pressure with norepinephrine does not improve microcirculatory blood flow: a prospective study. Crit Care 2009, 13(3):R92.

22. Boerma EC, Koopmans M, Konijn A et al.: Effects of nitroglycerin on sublingual microcirculatory blood flow in patients with severe sepsis/septic shock after a strict resuscitation protocol: A double-blind randomized placebo controlled trial. Crit Care Med 2010, 38(1):93-100.

23. Dubin A, Pozo MO, Casabella CA et al.: Comparison of 6% hydroxyethyl starch 130/0.4 and saline solution for resuscitation of the microcirculation during the early goal-directed therapy of septic patients. J Crit Care 2010, 25(4):659.e1-659.e8.

24. van Genderen M, Gommers D, Klijn E et al.: Postoperative sublingual microcirculatory derangement following esophagectomy is prevented with dobutamine. Clin Hemorheol Microcirc 2011, 48(4):275-283.

25. Ospina-Tascon G, Neves AP, Occhipinti G et al.: Effects of fluids on microvascular perfusion in patients with severe sepsis. Intensive Care Med 2010, 36(6):949-955.

26. Vincent JL, Bihari DJ, Suter PM et al.: The prevalence of nosocomial infection in intensive care units in Europe. Results of the European Prevalence of Infection in Intensive Care (EPIC) Study. EPIC International Advisory Committee. JAMA 1995, 274(8):639-644.

27. Vincent JL, Sakr Y, Sprung CL et al.: Sepsis in European intensive care units: results of the SOAP study. Crit Care Med 2006, 34(2):344-353.

28. Vellinga NAR, Boerma EC, Koopmans M et al.: Study Design of the Microcirculatory Shock Occurrence in Acutely Ill Patients (microSOAP): an International Multicenter Observational Study of Sublingual Microcirculatory Alterations in Intensive Care Patients. Critical Care Research and Practice 2012, 2012:1-7.

29. De Backer D, Hollenberg S, Boerma C et al.: How to evaluate the microcirculation: report of a round table conference. Crit Care 2007, 11(5):R101.

30. Boerma EC, Mathura KR, van der Voort PHJ et al.: Quantifying bedside-derived imaging of microcirculatory abnormalities in septic patients: a prospective validation study. Crit Care 2005, 9(6):R601-R606.

4


31. Klijn E, Den Uil CA, Bakker J et al.: The heterogeneity of the microcirculation in critical illness. Clin Chest Med 2008, 29(4):643-54, viii.

32. Dobbe JGG, Streekstra GJ, Atasever B et al.: Measurement of functional microcirculatory geometry and velocity distributions using automated image analysis. Med Biol Eng Comput 2008, 46(7):659-670.

33. Spanos A, Jhanji S, Vivian-Smith A et al.: Early microvascular changes in sepsis and severe sepsis. Shock 2010, 33(4):387-391.

34. Omar YG, Massey M, Andersen LW et al.: Sublingual microcirculation is impaired in post-cardiac arrest patients. Resuscitation 2013, 84(12):1717-1722.

35. Filbin MR, Hou PC, Massey M et al.: The Microcirculation Is Preserved in Emergency Department Low-acuity Sepsis Patients Without Hypotension. Acad Emerg Med 2014, 21(2):154-162.

36. Knaus WA, Draper EA, Wagner DP et al.: APACHE II: a severity of disease classification system. Crit Care Med 1985, 13(10):818-829.

37. Bakker J: Lactate: may I have your votes please? Intensive Care Med 2001, 27(1):6-11.

38. Shapiro NI, Howell MD, Talmor D et al.: Serum lactate as a predictor of mortality in emergency department patients with infection. Ann Emerg Med 2005, 45(5):524-528.

39. Nichol AD, Egi M, Pettila V et al.: Relative hyperlactatemia and hospital mortality in critically ill patients: a retrospective multi-centre study. Crit Care 2010, 14(1):R25.

40. Jansen TC, van Bommel J, Woodward R et al.: Association between blood lactate levels, Sequential Organ Failure Assessment subscores, and 28-day mortality during early and late intensive care unit stay: a retrospective observational study. Crit Care Med 2009, 37(8):2369-2374.

41. Parker MM, Shelhamer JH, Natanson C et al.: Serial cardiovascular variables in survivors and nonsurvivors of human septic shock: heart rate as an early predictor of prognosis. Crit Care Med 1987, 15(10):923-929.

42. Torgersen C, Meichtry J, Schmittinger CA et al.: Haemodynamic variables and functional outcome in hypothermic patients following out-of-hospital cardiac arrest. Resuscitation 2013, 84(6):798-804.

43. Han Z, Yan-min Y, Jun Z et al.: Prognostic value of admission heart rate in patients with ST-segment elevation myocardial infarction: role of type 2 diabetes mellitus. BMC Cardiovasc Disord 2012, 12:104.

44. Dünser MW, Ruokonen E, Pettilä V et al.: Association of arterial blood pressure and vasopressor load with septic shock mortality: a post hoc analysis of a multicenter trial. Crit Care 2009, 13(6):R181.

45. Hoke RS, Müller-Werdan U, Lautenschläger C et al.: Heart rate as an independent risk factor in patients with multiple organ dysfunction: a prospective, observational study. Clin Res Cardiol 2012, 101(2):139-147.

46. Smith I, Kumar P, Molloy S et al.: Base excess and lactate as prognostic indicators for patients admitted to intensive care. Intensive Care Med 2001, 27(1):74-83.

47. Magder SA: The ups and downs of heart rate. Crit Care Med 2012, 40(1):239-245.

| Chapter 478

48. Grander W, Müllauer K, Koller B et al.: Heart rate before ICU discharge: a simple and readily available predictor of short- and long-term mortality from critical illness. Clin Res Cardiol 2013, 102(8):599-606.

49. Schmittinger CA, Torgersen C, Luckner G et al.: Adverse cardiac events during catecholamine vasopressor therapy: a prospective observational study. Intensive Care Med 2012, 38(6):950-958.

50. Disegni E, Goldbourt U, Reicher-Reiss H et al.: The predictive value of admission heart rate on mortality in patients with acute myocardial infarction. SPRINT Study Group. Secondary Prevention Reinfarction Israeli Nifedipine Trial. J Clin Epidemiol 1995, 48(10):1197-1205.

51. Bone RC, Balk RA, Cerra FB et al.: Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine. Chest 1992, 101(6):1644-1655.

52. Sakr Y, Dubois MJ, De Backer D et al.: Persistent microcirculatory alterations are associated with organ failure and death in patients with septic shock. Crit Care Med 2004, 32(9):1825-1831.

53. Arnold RC, Parrillo JE, Phillip Dellinger R et al.: Point-of-care assessment of microvascular blood flow in critically ill patients. Intensive Care Med 2009, 35(10):1761-1766.

54. Boerma EC, van der Voort PHJ, Spronk PE et al.: Relationship between sublingual and intestinal microcirculatory perfusion in patients with abdominal sepsis. Crit Care Med 2007, 35(4):1055-1060.

55. Bezemer R, Bartels SA, Bakker J et al.: Clinical review: Clinical imaging of the sublingual microcirculation in the critically ill - where do we stand? Crit Care 2012, 16(3):224.

4


SUPPLEMENTAL CONTENT

This was provided as supplemental digital content to ‘International study on Microcirculatory Shock Occurrence in Acutely ill Patients (microSOAP)’ as accepted for publication in Critical Care Medicine

Multivariable logistic regression

Additional data Table S1: Patient characteristics Table S2: Patient characteristics, hospital survivors vs. hospital non-survivors Table S3: overview of hemodynamic monitoring Table S4: Patient characteristics of patients with and without tachycardiaSDF analysis

page 80

page 84page 86page 87page 88page 89

| Chapter 480

MULTIVARIABLE LOGISTIC REGRESSION

We applied multiple imputation analysis in order to correct for missing data. Twenty imputations were generated. APACHE, SOFA score (missing for respectively 16% and 18% of patients) were imputated based on age, gender, lactate level ≤/>1.5 mmol/L, stay in ICU before SDF imaging, heart rate and the use of mechanical ventilation and renal replacement therapy. SOFA scores were omitted during subsequent modelling because there was lower availability of this score as compared to APACHE score. Of note, some components of the SOFA score were included as potential independent variables.

Furthermore, we dichotomized cumulative vasopressor index and stay in ICU before SDF imaging (based on a non linear association of risk levels with outcome).

- Stay in ICU before SDF imaging ≤ 24 hours or > 24 hours

- Vasopressor use vs. no vasopressor use

Because lactate levels were available for 67% of patients and missing lactate level was due to absence of indications to perform this test, we decided not to use multiple imputation for lactate levels, but to divide patients in two groups: ‘Lactate levels ≤ 1.5 mEq/L or lactate unavailable’ vs. ‘Lactate > 1.5 mEq/L.

Odds ratios for mortality in patients without lactate measurements were similar to those in patients with lactate level < 1.5 mEq/L. Therefore, it was decided to group the patients without lactate measurements with patients with this low lactate level. Thus, this relative risk also applies in patients in whom lactate will not be assessed. See table below for odds ratios with lactate groups as categorical variables.

OR (95% CI) PNo lactate available (reference category) 1

Lactate ≤ 1.5 mEq/L 1.07 (0.65-1.76) 0.798

Lactate > 1.5 mEq/L 2.90 (1.74-4.84) <0.001

For admission diagnoses, data were re-grouped in a similar way. The seven groups (surgery, sepsis, cardiac disease, neurological disorders, trauma, respiratory insufficiency, other) as presented in the baseline table were considered too fragmented to include in the multivariable analysis. Therefore we regrouped the admission diagnosis based on their odds ratios.

4


Based on the original division in seven groups:

P OR 95%-CIOverall effect (Surgery = reference group, n=163) .000 1

Sepsis (n=84) .000 3.572 1.939-6.580

Trauma (n=49) .157 1.786 0.801-3.984

Hemorrhage (n=10) .170 2.760 0.647-11.765

Neurological disorders (n=55) .082 1.932 0.919-4.061

Cardiac disease (n=56) .001 3.409 1.708-6.804

Respiratory insufficiency (n=42) .000 6.072 2.896-12.732

Other (n=42) .075 2.094 0.927-4.727

Regrouping:

P OR 95%-CIOverall effect (surgery = reference group, n=163) .000 1

Sepsis (n=84) .000 3.572 1.939-6.580

Trauma/hemorrhage/other (n=101) .030 2.000 1.070-3.737

Respiratory/cardiac disease (n=98) .000 4.396 2.449-7.888

Neurological disorders (n=55) .082 1.932 0.919-4.061

The following a priori defined potential variables were tested in univariable logistic regression with hospital mortality as outcome variable (including first order interactions):

APACHE II score (P<0.001)length of stay (LOS) in the ICU prior to SDF imaging > 24 hours (P <0.001) admission diagnosis (5 groups: surgery-sepsis-cardiac disease etc.): overall effect in univariable analysis P<0.001hemoglobin (P = 0.004)arterial lactate > 1.5 mEql/L (P<0.001)tachycardia (heart rate >90 beats per minute [bpm]) (P < 0.001)mean arterial pressure (MAP) (P<0.001)fluid balance (P=0.802)vasopressor use (P<0.001)small vessel MFI (P=0.788)the presence of an abnormal MFI (P = 0.108)proportion of perfused small vessels (PPV) (P=0.066)perfused small vessel density (PVD) (P=0.579)total small vessel density (TVD) (P=0.967)renal replacement therapy (P <0.001)mechanical ventilation (P < 0.001)

| Chapter 482

Because of a significant interaction between tachycardia and both small vessel MFI (P=0.030) and abnormal MFI (P=0.012), subgroup analysis as described in the manuscript is an appropriate step in order to take the effect of this interaction into account.

In patients with tachycardia, these were the results of univariable logistic regression with hospital mortality as outcome variable:

APACHE II score (P=0.003)length of stay (LOS) in the ICU prior to SDF imaging > 24 hours (P = 0.011) admission diagnosis (5 groups: surgery-sepsis-cardiac disease etc.): overall effect in univariable analysis P=0.003hemoglobin (P = 0.639)arterial lactate > 1.5 mmol/L (P<0.001)mean arterial pressure (MAP) (P=0.019)fluid balance (P=0.742)vasopressor use (P<0.001)small vessel MFI (P=0.047)the presence of an abnormal MFI (P = 0.003)proportion of perfused small vessels (PPV) (P=0.082)perfused small vessel density (PVD) (P=0.746)total small vessel density (TVD) (P=0.899)renal replacement therapy (P=0.008)mechanical ventilation (P=0.005)

The above mentioned variables with P< 0.25 in univariable analysis were included in backwards stepwise modelling.

With every step, the variables with the highest P-value was omitted, until all variables in the model had P-values below 0.05.

4


The following models evolved from the analysis.

A. Hospital mortality OR (95% CI) PAPACHE II 1.05 (1.02-1.08) 0.001


Lactate level > 1.5 mEq/l* 2.15 (1.28-3.62) 0.004

Heart rate > 90 bpm 2.71 (1.67-4.39) <0.001

MAP (mmHg) 0.979 (0.963-0.996) 0.013


Renal replacement therapy 2.26 (1.07-4.80) 0.034



Sepsis 2.07 (1.02-4.22) 0.045


Resp. insufficiency/cardiac disease 3.44 (1.73-6.87) <0.001


B. Hospital mortality for HR>90 OR (95% CI) PLactate level > 1.5 mEq/l* 2.84 (1.36-5.92) 0.005


Abnormal MFI 3.24 (1.30-8.06) 0.011




Sepsis 3.08 (1.27-7.45) 0.013


Resp. insufficiency/cardiac disease 3.54 (1.38-9.08) 0.008


A. Multivariable logistic regression for hospital mortality (all patients). Average Nagelkerke R2 = 0.37 (range: 0.36-0.41) (P<0.001), average Hosmer & Lemeshow Chi-square 5.303 (range 1.383-9.400), P=0.710 (range 0.310-0.994) . B. Multivariable logistic regression for hospital mortality for patients with heart rate > 90/min (n= 204). Nagelkerke R2 = 0.32 (P<0.001), Hosmer & Lemeshow Chi-square 5.576, P=0.59. P=0.022 for overall effect of admission diagnosis. Because APACHE II score was not included in this model, models in all imputations were equal. OR = odds ratio, 95% CI = 95% confidence interval, ICU = intensive care unit, SDF = Sidestream Dark Field imaging, bpm = beats per minute, MAP = mean arterial pressure, abnormal MFI = MFI<2.6 for vessels < 20 µm.*As compared to patients with a lactate <1.5 mEq/L or no lactate measurement available. ** As compared to patients with surgery as admission diagnosis

| Chapter 484

ADDITIONAL DATA

Table S1 (Supplemental Digital Content). Patient characteristics

Age (years) 62 [51-73]

Male sex – no. (%) 297 (59%)

APACHE IIa 15 [10-21]

SOFAb 5 [2-8]

Time in ICU before SDF imaging (days) 3.0 [0.86-8.0]

Length of stay in ICU (days) 9 [3-20]

Length of stay in hospital (days) 18 [9-33]

ICU mortality – no. (%) 108 (21.6)

In-hospital mortality – no. (%) 140 (28.4)

Reason for ICU admission

Surgery – no. (%) 163 (33)

Cardiac surgery– no. (% of surgery) 80 (49)

Non-cardiac surgery– no. (% of surgery) 83 (51)

Sepsis – no. (%) 84 (17)

Abdominal sepsis– no. (% of sepsis) 30 (36)

Pneumosepsis– no. (% of sepsis) 32 (38)

Other– no. (% of sepsis) 22 (26)

Cardiac disease – no. (%) 56 (11)

Neurological disorders – no. (%) 55 (11)

Trauma – no. (%) 49 (9.8)

Respiratory insufficiency – no. (%) 42 (8.4)

Other – no. (%) 52 (10)

Comorbidity

Diabetes mellitus – no. (%) 84 (16.8)

Heart failure – no. (%) 72 (14.4)

Malignancy – no. (%) 50 (10.0)

Chronic obstructive pulmonary disease – no. (%) 52 (10.4)

Heart rate (beats per minute) 88 ± 19

Mean arterial pressure (mmHg) 84 [75-94]

Hemoglobin (g/dL) 10.2 [8.7-11.6]

Central venous pressure (mmHg) 10 [7-14]

Arterial lactate (mEq/L) 1.2 [0.9-2.0]

Fluid balance (24 hours before SDF) (mL) 1000 [0-2319]

Vasopressive drugs

Patients treated – no. (%) 154 (31)

Cumulative vasopressor indexc 3.3 ± 1.6

Mechanical ventilation – no. (%) 266 (53)

4


Renal replacement therapy – no. (%) 44 (8.8)

Abnormal microcirculation (n, %) 83 (17)

MFI small vessels (AU) 2.93 [2.67-3.00]

MFI large vessels (AU) 3.00 [2.92-3.00]

TVD, mm/mm2 (small vessels) 19.26 ± 3.98

PVD, mm/mm2 (small vessels) 18.56 ± 3.93

PPV (small vessels) 0.98 [0.95-0.99]

De Backer score (small vessels, n/mm) 11.71 ± 2.48

De Backer score (perfused small vessels, n/mm) 11.29 ± 2.44

Heterogeneity Index (small vessels) 0.34 [0.00-0.37]

Values are mean ± standard deviation or median [interquartile range] unless specified otherwise.aScores on the Acute Physiologic and Chronic Health Evaluation II (APACHE II) scale range from 0 to 71 (higher values indicate more severe disease). APACHE II was calculated for the first 24 hours of ICU admission.bScores on the Sequential Organ Failure Assessment (SOFA) scale range from 0 to 4 for each organ system, with higher scores indicating more severe organ dysfunction. cTrzeciak et al. Intensive Care Med 2008;34:2210-7 Abnormal microcirculation, small vessel MFI < 2.6. MFI, microvascular flow index. Cutoff value for small vessels: < 20 µm. TVD, total vessel density. PVD, perfused vessel density. PPV, proportion of perfused vessels. Follow-up data on mortality and length of stay were available for >98% of patients.

Table S1 (Supplemental Digital Content). Patient characteristics. Continued

| Chapter 486

Table S2 (Supplemental Digital Content). Patient characteristics, hospital survivors vs. hospital non-survivors

Hospital-survivors N = 353 (72%)

Hospital non-survivors N = 140 (28%)

P

Age (years) 61 [51-72] 66 [54-76] 0.007*

Male sex – no. (%) 217 (61) 74 (53) 0.09

APACHE IIa 13 [9-18] 21 [16-26] <0.001*

SOFAb 3 [2-6] 8 [6-11] <0.001*

Time in ICU before inclusion (days) 2.0 [0.67-7.1] 5.0 [2.0-11.0] <0.001*

Reason for ICU admission <0.001*

Surgery – no. (%) 138 (39) 25 (18)

Sepsis – no. (%) 51 (14) 33 (24)

Cardiac disease – no. (%) 34 (9.6) 21 (15)

Respiratory insufficiency – no. (%) 20 (5.7) 22 (16)

Neurological disorders – no. (%) 40 (11) 14 (10)

Trauma – no. (%) 34 (9.6) 11 (7.9)

Haemorrhage – no (%) 6 (1.7) 3 (2.1)

Other – no. (%) 30 (8.5) 11 (7.9)

Comorbidity 0.05

Diabetes mellitus – no. (%) 54 (15) 30 (21)

Heart failure – no. (%) 41 (12) 31 (22)

Malignancy – no. (%) 30 (8.5) 20 (14)

Chronic obstructive pulmonary disease – no. (%)

37 (10) 15 (11)

Heart rate (beats per minute) 84 ± 17 97 ± 21 <0.001*

Heart rate > 90 beats per minute – no. (%)

113 (32) 84 (60) <0.001*

Mean arterial pressure (mmHg) 87 ± 15 81 ± 15 <0.001*

Hemoglobin (g/dL) 10.3 [10.3-11.9] 9.7 [8.4-11.0] <0.001*

Arterial lactate (mEq/L) 1.20 [0.90-1.80] 1.69 [1.00-2.53] <0.001*

Fluid balance (24 hours before SDF) (mL)

888 [-35 – 2350] 1118 [63-2344] 0.27

Vasopressive drugs

Patients treated – no. (%) 82 (23) 71 (51) <0.001*

Cumulative vasopressor indexc 3.1 ± 1.5 3.5 ± 1.7 0.19

Mechanical ventilation – no. (%) 162 (46) 103 (74) <0.001*

Renal replacement therapy – no. (%) 15 (4.2) 28 (20) <0.001*

Abnormal microcirculation (n, %) 52 (15) 29 (21) 0.11

MFI small vessels (AU) 2.93 [2.67-3.00] 2.93 [2.67-3.00] 0.91

MFI large vessels (AU) 3.00 [2.92-3.00] 3.00 [2.90-3.00] 0.74

4


Hospital-survivors N = 353 (72%)

Hospital non-survivors N = 140 (28%)

P

TVD, mm/mm2 (small vessels) 19.24 ± 4.00 19.22 ± 3.92 0.97

PVD, mm/mm2 (small vessels) 18.61 ± 3.95 18.39 ± 3.89 0.58

PPV (small vessels) 0.98 [0.95-0.99] 0.97 [0.94-0.99] 0.092

De Backer score (small vessels, n/mm) 11.69 ± 2.47 11.71 ± 2.59 0.94

De Backer score (perfused small vessels, n/mm)

11.30 ± 2.43 11.21 ± 2.47 0.71

Heterogeneity Index (small vessels) 0.34 [0.00-0.37] 0.34 [0.00-0.37] 0.86

Values are mean ± standard deviation or median [interquartile range] unless specified otherwise.Mann-Whitney U test for non-normally distributed parameters. T-test for normally distributed variables. Fisher’s exact test for dichotomous variables. *A two-sided P<0.05 was considered statistically significant.aScores on the Acute Physiologic and Chronic Health Evaluation II (APACHE II) scale range from 0 to 71 (higher values indicate more severe disease). APACHE II was calculated for the first 24 hours of ICU admission. bScores on the Sequential Organ Failure Assessment (SOFA) scale range from 0 to 4 for each organ system, with higher scores indicating more severe organ dysfunction. cTrzeciak et al. Intensive Care Med 2008;34:2210-7 Abnormal microcirculation, small vessel MFI < 2.6. MFI, microvascular flow index. Cutoff value for small vessels: < 20 µm. TVD, total vessel density. PVD, perfused vessel density. PPV, proportion of perfused vessels.

Central venous pressure was obtained in 40% of patients (10 [7-14] mmHg), and cardiac index was available for 6.2% of patients at the time of SDF imaging (3.2 [2.4-4.0] L/min/m2). Mixed venous or central venous saturation (S(c)vO2) was obtained in 20% of patients (72% [67-77%]).

Table S3. overview of hemodynamic monitoring

Hemodynamic monitoring N (%)Central venous pressure 198 (40)

Central venous oxygen saturation 86 (17)

Pulmonary artery catheter 21 (4.2)

Pulse contour analysis 13 (2.6)

Oesophagus Doppler 2 (0.4)

Table S2 (Supplemental Digital Content). Patient characteristics, hospital survivors vs. hospital non-survivors. Continued

| Chapter 488

Table S4. Patient characteristics of patients with and without tachycardia

HR ≤ 90 bpm (N = 297 (59%))

HR > 90 bpm (N = 204 (41%))

P

Age (years) 63 [52-74] 61 [49-72] 0.115

Male sex – no. (%) 183 (62) 112 (55) 0.165

APACHE IIa 14 [9-20] 17 [11-23] <0.001

SOFAb 4 [2-7] 6 [3-9] <0.001

Time in ICU before SDF imaging (days) 2.0 [0.67-7.1] 5.0 [2.0-10] <0.001

ICU mortality – no. (%) 48 (16) 60 (29) <0.001

Hospital mortality – no. (%) 56 (19) 84 (41) <0.001

Length of stay in ICU (days) 8 [2-17] 12 [5-23] <0.001

Length of stay in hospital (days) 15 [8-29] 21 [12-41] <0.001

Reason for ICU admission 0.155

Surgery – no. (%) 95 (32) 67 (33)

Sepsis – no. (%) 39 (13) 45 (22)

Cardiac disease – no. (%) 35 (12) 20 (9.8)

Respiratory insufficiency – no. (%) 22 (7.4) 20 (9.8)

Neurological disorders – no. (%) 40 (13) 15 (7.4)

Trauma – no. (%) 30 (10) 19 (9.3)

Other – no. (%) 36 (12) 18 (8.8)

Comorbidity 0.911

Diabetes mellitus – no. (%) 49 35

Heart failure – no. (%) 41 30

Malignancy – no. (%) 24 26

Chronic obstructive pulmonary disease – no. (%)

33 19

Heart rate (beats per minute) 75 ± 10 107 ± 14 <0.001

Mean arterial pressure (mmHg) 85 ± 16 84 ± 16 0.468

Hemoglobin (mmol/L) 6.5 [5.7-7.7] 6.0 [5.3-6.7] <0.001

Arterial lactate (mmol/L) 1.2 [0.9-2.0] 1.3 [0.9-2.1] 0.282

Fluid balance (24 hours before SDF) (mL) 928 [-35-2100] 1145 [50-2747] 0.132

Vasopressive drugs

Patients treated – no. (%) 78 (26) 76 (37) 0.007

Cumulative vasopressor indexc 3.4 ±1.4 3.2 ± 1.8 0.570

Mechanical ventilation – no. (%) 143 123 0.004

Renal replacement therapy – no. (%) 19 25 0.018

Values are mean ± standard deviation or median [interquartile range] unless specified otherwise.Mann-Whitney U test for non-normally distributed parameters. T-test for normally distributed variables. Chi-square test for dichotomous variables. *A two-sided P<0.05 was considered statistically significant.aScores on the Acute Physiologic and Chronic Health Evaluation II (APACHE II) scale range from 0 to 71, with higher values indicating more severe disease.bScores on the Sequential Organ Failure Assessment (SOFA) scale range from 0 to 4 for each organ system, with higher scores indicating more severe organ dysfunction.

4


SDF analysis

HR ≤ 90 bpm (N = 297 (59%))

HR > 90 bpm (N = 204 (41%))

P

Abnormal microcirculation (n, %) 51 (17) 32 (16) 0.695

MFI small vessels (arbitrary units) 2.93 [2.67-3.00] 2.93 [2.73-3.00] 0.110

MFI large vessels (arbitrary units) 3.00 [2.90-3.00] 3.00 [2.93-3.00] 0.415

TVD, mm/mm2 (small vessels) 19.42 ± 3.80 19.01 ± 4.26 0.262

PVD, mm/mm2 (small vessels) 18.71 ± 3.80 18.35 ± 4.15 0.317

PPV (small vessels) 0.98 [0.95-0.99] 0.97 [0.95-0.99] 0.421

De Backer score (small vessels) 11.76 ± 2.39 11.65 ± 2.63 0.646

De Backer score (perfused small vessels) 11.32 ± 2.37 11.24 ± 2.55 0.691

Heterogeneity Index (small vessels) 0.34 [0.00-0.37] 0.34 [0.00-0.37] 0.297

Values are mean ± standard deviation or median [interquartile range] unless specified otherwise. Mann-Whitney U test for non-normally distributed parameters. T-test for normally distributed variables. *A two-sided P<0.05 was considered statistically significantAbnormal microcirculation, small vessel MFI < 2.6. MFI, microvascular flow index. Cutoff value for small vessels: < 20 µm. TVD, total vessel density. PVD, perfused vessel density. PPV, proportion of perfused vessels.

uva-dare (digital academic repository) microcirculatory ... · 16. derriford hospital, plymouth...

Documents