Variability of microcirculatory measurements in healthy volunteers
Abstract
Reliable assessment of the microcirculation is important to investigate microcirculatory properties in various disease states. The GlycoCheck system automatically analyzes sublingual sidestream dark field images to determine the perfused boundary region (PBR; a measure of glycocalyx thickness), red blood cell filling percentage, and microvascular vessel density. Although GlycoCheck has been used to study the microcirculation in patients, little is known about the reproducibility of measurements in healthy volunteers. We assessed intra- and interobserver agreement by having two experienced observers perform three consecutive microcirculation measurements with the GlycoCheck system in 49 healthy volunteers. Intraobserver agreement of single measurements were poor (intraclass correlation coefficients (ICCs) < 0.4) for PBR, red blood cell filling percentage and microvascular vessel density. ICCs increased to values > 0.6 (indicating good reproducibility) for all parameters when performing and averaging three consecutive measurements. No systematic differences were observed between observers for any parameter. Interobserver variability was fair for PBR (ICC = 0.53) and red blood cell filling percentage (ICC = 0.58) and poor for perfused vessel density (ICC = 0.20). In conclusion, GlycoCheck software can be used with acceptable reliability and reproducibility for microcirculation measurements on a population level when averaging three consecutive measurements. Repeated measurements are preferably performed by the same observer.
Introduction
Actual exchange of oxygen, nutrients and waste products takes place in the microcirculation, consisting of blood vessels of < 250 μm in diameter (microvessels)1. The glycocalyx lines the luminal side of the microvessels and forms a semi-permeable barrier between blood and endothelium2. It plays an important role in the regulation of coagulation factors, prevention of leakage of plasma components, activation/inhibition of platelets and adhesion blocking of leukocytes3,4. The role of the microcirculation and specifically of the glycocalyx in the pathophysiology of different disease states is increasingly acknowledged5,6.
Reliable assessment of the microcirculation and its pathology is important to investigate its properties in various disease states5,6. Eventually, this may lead to the implementation of microcirculatory monitoring as an adjunct to guide therapy. Microcirculation can be assessed non-invasively and in vivo using one of the following video microscopy techniques: orthogonal polarized spectral imaging, side-stream dark field (SDF) imaging, and incident dark field (IDF) imaging7. The application of these video microscopes is currently limited to scientific research due to several factors. First of all, knowledge on reference values in different populations is limited. Secondly, validated microcirculatory measurement targets that can influence clinical outcome are lacking6,8,9. To develop such targets and move towards the implementation of microcirculatory-guided therapy in clinical practice, a validated, readily applicable and reliable assessment tool is imperative. Most camera systems, however, rely on an off-line, manual analysis of the captured videos to determine microcirculatory variables. This is time consuming and requires thoroughly trained staff.
GlycoCheck (Microvascular Health Solutions Inc., Salt Lake City, UT, USA) is a commercially available software package that automates assessment of the microcirculation based on SDF. It selects and analyses SDF images that are of sufficient quality (adequate focus, adequate contrast and limited movement) and determines glycocalyx thickness, microvascular vessel density and red blood cell filling percentage10. Unfortunately, inter- and intraobserver agreement of this technique when used to assess the microcirculation in pathological conditions, is poor11,12,13. Averaging three consecutive measurements was shown to achieve intraobserver variability that can be classified as “excellent” in pathological conditions such as critical illness and smoking11,12. Whether this is due to the biological variability caused by pathological conditions or due to the inherent inaccuracy of GlycoCheck system is yet to be determined. The aim of the present study was to assess the reproducibility of GlycoCheck measurements in young, healthy volunteers without any pathological condition that affects microcirculation. To this effect, we evaluated both intra- and interobserver agreement.
Methods
We included non-smoking, healthy volunteers between 18 and 40 years of age. Exclusion criteria were diabetes mellitus type 1 or 2, nicotine-use and chronic illness for which chronic medication is used. Additionally, volunteers were excluded in case of oral bleeding, oral wounds and oral infections as these reduce the measurement quality. We recorded age, sex, height and weight of the volunteers.
Measurements
Measurements were performed using an SDF camera (CapiScope HVCS, KK Technology, Honiton, UK) fitted with GlycoCheck software (Microvascular Health Solutions Inc., Salt Lake City, UT, USA). Two researchers (MEB and BB), both experienced in performing sublingual measurements with this tool, performed three consecutive measurements for each subject. The measurement order of the two researchers was randomly determined. Measurements were taken with subjects in supine position with the researcher standing behind the headboard. Volunteers were asked to swallow any saliva prior to measuring after which the camera was manually placed and held still in the sublingual region. To limit movement of the camera, the researchers could rest their wrist on the lower jaw of the volunteers. Care was taken to limit pressure artifacts by ensuring erythrocytes could be seen traveling through the blood vessels during the measurements. Images were recorded only in the absence of air bubbles, excessive amounts of saliva, excessive loops of the vessels and excessive amounts of large venules. All measurements in an individual volunteer were performed within 30 min.
GlycoCheck parameters
The GlycoCheck software records movies of 1 s that consist of 23 frames. Recording is initiated automatically when the software deems the images of sufficient quality, meaning that the intensity and focus are sufficient for calculations and that the camera is held sufficiently still. Vessels are automatically detected and measurement points are defined at 10 µm intervals. GlycoCheck limits its calculations to vessels with a width between 5 and 25 µm. A measurement is complete when 3000 measurement points have been acquired.
Perfused boundary region
The inner layer of the glycocalyx is penetrable to red blood cells and, hence, also called the perfused boundary region (PBR)10. An intact glycocalyx is thicker and less penetrable to red blood cells, leading to a thinner PBR, compared to a damaged glycocalyx. The PBR is thus an inverse measure of the glycocalyx thickness. GlycoCheck software calculates the PBR from the intensity profile at every measurement point10. A more gradual increase of the intensity profile means a thicker PBR – indicating a thinner glycocalyx.
Red blood cell filling percentage
The red blood cell filling percentage, a measure of microvascular perfusion, is defined as the median percentage of time that red blood cells are present at each measurement point10.
Microvascular vessel density
GlycoCheck calculates microvascular vessel density from the amount of measurement points—as each measurement point represents 10 µm of microvessel length10. Cumulative microvessel length in µm was thus equal to the amount of measurement points multiplied by 10. Microvascular vessel density in µm/mm2 was calculated by dividing the cumulative microvessel length by the total recorded area in mm2.
Statistics
Data are presented as median [25–75th percentile] or as mean ± standard deviation (SD) as appropriate. Normality was assessed using the Kolmogorov–Smirnov test and the Shapiro Wilk tests for normality. Data analysis was performed with SPSS (version 27; IBM, Armonk, NY, USA). P-values < 0.05 were considered to be statistically significant. Volunteers were excluded from the analysis in case of a missing measurement.
Intraobserver agreement for single measurements was assessed with Intraclass Correlation Coefficients (ICCs) by means of a two-way random model with absolute agreement (type ICC(2,1) according to the Shrout and Fleiss convention14) and reported as ICC (95% confidence interval; CI) for each of the two observers. For the mean of three measurements ICCs were assessed by means of a two-way random model with absolute agreement for the mean of multiple measurements (type ICC(2, k); where k = 3). ICCs were deemed as poor (< 0.40), fair (0.40–0.60), good (0.60–0.75) or excellent (> 0.75) according to the guidelines written by Cicchetti15.
Sample size calculation was performed such that an ICC of 0.6 or higher (indicating good reproducibility) would be detected with a 95% confidence interval width of at most 0.3. Based on the method described by Lew et al., we calculated a sample size of 45 volunteers16.
The average values of the 3 measurements taken by each observer were used for interobserver agreement analysis. Interobserver agreement was assessed using Bland–Altman plots and with ICCs of a two-way random model with absolute agreement for single measurements (type ICC (2,1)). The paired t-test was used to assess whether there was a systematic difference between both observers.
Ethics
The protocol was reviewed and approved by our institutional Review Board (METC: #2017-0122). Volunteers were informed about the study in word and writing. Written informed consent was obtained from all participants. The study was performed in accordance with the Declaration of Helsinki.
Results
A total of 49 volunteers participated in this prospective study. Table 1 shows volunteer characteristics. Forty-six of the volunteers completed all 6 measurements (three for each observer). Three volunteers had one measurement less due to technical problems with the GlycoCheck.