Short-term effects of passive mobilization on the sublingual microcirculation and on the systemic circulation in patients with septic shock

We included patients if they met all the following criteria: 18 years of age or older, septic shock [27] with noradrenaline requirement between 0.1 and 1.0 mcg/kg/min, sedation with a Ramsay scale of 6 [28], mechanical ventilation, an arterial line for blood pressure monitoring, and an indwelling subclavian or jugular central venous catheter. All patients were included after the resuscitation phase of shock.

The exclusion criteria were intracranial hypertension, status epilepticus, or cardiac arrhythmia at the time of the study; acute myocardial infarction (Killip classification 4); thrombocytopenia (<30,000 cells/µL); hemoglobin <7.0 g/dL; leukocytosis (>50,000 cells/µL), peripheral arterial disease, worsening oxygenation status in the last 6 h (oxygen partial pressure to oxygen inspired fraction ratio decrease of >50), circulatory shock of multiple causes, or pregnant and moribund patients. We also excluded patients with medical conditions that prevented the use of a mobilization protocol, such as bone tumors, unconsolidated or unstable bone fracture, amputation, postoperative period for placement or removal of upper and lower limbs bone prosthesis, deep vein thrombosis or phlebitis, musculoskeletal deformity, and compartment syndrome in the upper or lower limbs. In addition, we excluded patients with mucosal injury or recent maxillofacial surgery that interfered with imaging acquisition.

Patients were kept in a semi-recumbent position (30°). Passive movements of the upper limbs (flexion–extension of the wrist, elbow, and shoulder) and lower limbs (flexion–extension of the ankle, knee, and hip) were performed in a counter-clockwise direction starting from the upper right. The frequency of 30 movements per minute was assured by a digital metronome MA-1 BLBK^® (KORG, Tokyo, Japan) with 5 min for each limb, totaling 20 min of exercise. One of the investigators (TTP), a trained physiotherapist, applied the passive mobilization protocol. Although this protocol is a standard of care in our unit, it is not routinely applied in those severely ill patients.

We observed all patients for 20 min before passive movement to assure that there were no significant variations in sedative and vasoactive medication infusions as well as hemodynamic and respiratory parameters. Before the intervention, the vasopressor dose was titrated by the attending physician targeting a mean arterial pressure of at least 65 mmHg. The protocol was interrupted if any of the following adverse events occurred: detection of active movement, hypoxemia (pulse oximetry <88%), hemodynamic instability (heart rate (HR) <45 bpm, mean arterial pressure (MAP) <60 mmHg or >120 mmHg), and changes in the doses of vasoactive or sedation drugs.

We recorded systemic hemodynamics and sublingual microcirculatory variables before (T0) and up to 10 min after (T1) passive movements. We assessed the sublingual microcirculation using sidestream dark field (SDF) imaging (Microscan^®; MicroVision Medical, Amsterdam, Netherlands). To ensure image quality and adequate reproducibility, we adopted recommended techniques for video acquisition [29]. After removing saliva and oral secretions, the probe was applied over the mucosa. Special care was taken to avoid pressure artifacts, which was verified by checking ongoing flow in larger microvessels. We collected three high-quality steady images of at least 10 s in different sublingual regions using a portable computer and an analog-to-digital video converter (ADVC110, Canopus Co., San Jose, CA, USA). The image quality was checked before being stored at full size as DV-AVI files. Digital sequences were stored and we used the AVA 3.0^® software (MicroVision Medical, Amsterdam, Netherlands) for automated vascular analysis. The correct procedure for vascular video analysis with AVA 3.0^® software includes some manual analysis steps, including erasing incorrectly detected vessel segments and adding undetected vessel segments. Two investigators (TTP and ATB) analyzed the images off-line. The images were presented in a random order to prevent image pairing. Microcirculatory parameters included the microcirculatory flow index (MFI), the total vascular density (TVD), the De Backer score (DBS), the perfused vascular density (PVD), the proportion of perfused vessels (PPV), and the heterogeneity index (HI) [29]. To calculate the MFI score, the image was divided into four quadrants and the predominant type of flow was assessed in each quadrant using semiquantitative criteria and scored as continuous (continuous flow for at least 10 s = 3), sluggish (slow but continuous flow = 2), intermittent (no flow ≤ 50% of time = 1), or absent (no flow ≥ 50% of time = 0). The MFI score averaged the values of the four quadrants. The TVD was calculated as the length of vessels divided by total area of the image. After drawing three equidistant horizontal and three equidistant vertical lines on the screen, we calculated the De Backer score as the number of vessels crossing the lines divided by the total length of the lines. The PPV (%) can be calculated as follows: 100 × (total number of vessels − [no flow + intermittent flow])/total number of vessels. The PVD was calculated as the length of perfused vessels (flow scores of 2–3) divided by the total area of the image. We used the values generated by AVA software report. As microcirculation is heterogeneous, we calculated the HI defined as the difference between maximal and minimal PPV divided by the mean PPV value of all sublingual sites. All the variables were related to vessels with diameters less than 20 µm.

The cardiac index (CI) was measured by transpulmonary thermodilution (VolumeView/EV1000™ system, Edwards Lifesciences, Irvine, CA, USA), pulmonary thermodilution using the traditional bolus method (Edwards Lifesciences LLC, Irvine, CA, USA), or echocardiography (SonoAce R7 device, Samsung Medison, Seoul, Korea) in patients without invasive monitoring. We also collected data on age, gender, body mass index, severity scores such as Sequential Organ Failure Assessment (SOFA) and Simplified Acute Physiologic Score (SAPS) 3, admission diagnosis, Charlson index, comorbidities, source of infection, sedative drugs, vasoactive drugs, time on vasopressors, mechanical ventilation parameters, ICU length of stay, and ICU mortality.

The data are expressed as a number (percent), the mean ± standard deviation, or the median (interquartile range), as appropriate. Continuous data were tested for normality by Shapiro–Wilk’s test. The systemic hemodynamic and microcirculatory variables were compared at T0 and T1 using a paired Wilcoxon test or paired t test, as appropriate. For the systemic hemodynamic variables, we defined post hoc as clinically relevant those changes greater than 10% using as reference the baseline [30]. As our primary hypothesis was that passive mobilization would potentially cause a worsening in the microcirculation we also hypothesize that this would occur mainly in patients using higher doses of noradrenaline. Thus, we analyzed our results according to the baseline noradrenaline dose. We classified as low-dose group dose those patients using less than 0.3 mcg/kg/min and as high-dose group those using ≥0.3 mcg/kg/min. We compared the microcirculatory variables at moments T0 and T1 within the groups, using t test or Wilcoxon test, as appropriate. We also measured the intensity of variation between T0 and T1 and compare the results of all microcirculatory variables between the groups using Mann–Whitney test or t test, as appropriate. The correlation between induced changes in systemic hemodynamics and microcirculatory variables was assessed using a Spearman correlation test.