Basic data
Table 2 shows the period of the survey points and the respective mobilisation method as well as the absolute and relative frequencies of the observations, indicating the cases excluded due to missing inclusion criteria.
(Table 2)
All in all, the mobilisations observed were performed exclusively by nurses (57.6% female, 42.4% male). Physiotherapists did not participate in the observed mobilisations. 5 observations were excluded at each of the survey points T1 and T2 due to the inclusion criterion “specialist training in anaesthesia and intensive care” or “nurse with at least 3 years of experience in intensive care” since the users did not meet these criteria (see also table 2).
The observations took place more frequently in the afternoon shift (65.7%) than in the morning shift at both survey points in terms of time of mobilisation.
Socio-demographic data of the users
The relative proportion of nurses with specialist training (PD with FWB) was 56.6%, and the overall average age was 33.1 years (±24.9/11.1; SD=6.89). The nurses without further training were younger on average ( =32.0 ±13.0/12.0; SD=5.5) and had a shorter intensive care experience ( =6.2 ± 13.8/3.0; SD=3.6) than those with further training.
Table 3 reflects the socio-demographic data of the users on specialist training, age, and intensive care experience, differentiated by time of survey and gender.
(Table 3)
Socio-demographic data of the patients
52.5% female patients and 47.5% male patients were included in the observation.
Normal-weight patients represented a higher proportion of the population overall (55.6%) and also represented the largest group at the respective observation points T1 (n=20: 57.1%) and T2/3 (n=35: 54.7%). There were more underweight people during robot-assisted mobilisation (n=13, 20.3%) than during conventional mobilisation (n=4, 11.4%). 11 (17.2%) overweight patients were mobilised with the support of the VEMOTION®, as compared to 8 (22.9%) patients with a BMI between 25.0 - 29.9 kg/m2 during conventional mobilisation, at the survey points T2/3. The overall mean BMI was 22.87 kg/m2 and not significantly different in T1 and T2/3 (T1 n=35, T2/3 n=64 U=972.0, z=-1.092, p=0.275).
With regard to medication, only the documentation of medication for analgesia, catecholamines, and simultaneous administration of both groups of medication was observed. 64 patients (62.5%) were analgosedated at T2/3, while only 4 patients (11.4%) were analgosedated during conventional mobilisation. Just under half of the patients (n= 31, 48.5%) mobilised with the VEMOTION® received circulatory support with catecholamines. This was necessary for only 13 patients at the survey point T1 (37.1%). During conventional mobilisation, however, more than half of the patients (n= 18, 51.4%) did not receive any analgosedation or catecholamine. At the time of robot-assisted mobilisations, only 15 patients (23.4%) were neither analgosedated nor did they need any catecholamine.
11 (84.6%) of the 13 invasively ventilated patients in T1 had a tracheostomy and only 2 (15.4%) had a tube. During robot-assisted mobilisation (T2/3), significantly more of the 36 invasively ventilated patients (77.8%, n=28) were fitted with a tube than with a tracheostomy.
Table 4 shows the absolute and relative distribution of the variables ventilation mode and ventilation access differentiated by T1 and T2/3.
(Table 4)
There were a total of 28 mentions of ingoing and outgoing tubes in T1 and 103 in T2/3. Table 5 shows the absolute frequencies sorted by drains, ingoing catheters, and extracorporeal therapy devices.
(Table 5)
Work organisation
The process-related data on the organisation of work were documented based on the time required in the process steps of preparation, execution, follow-up, and the calculated total duration of mobilisation, the number of specialist staff mobilising, and the aids used as well as the application of kinaesthesia.
Significant differences were evident in the time required between conventional and robot-assisted mobilisation in all sub-steps of mobilisation. Figure 3 shows the median time required in minutes, differentiated by preparation, execution, follow-up, and total duration of mobilisation.
The Mann-Whitney-U test was used to show significant and high effect sizes (in accordance with Cohen) (23) between the times required for conventional and robot-assisted mobilisation. The greatest difference between T1 and T2/3 was seen in the preparation time (see table 6).
(Table 6)
A bed sheet was used as a support during conventional mobilisation (n=7), and a slide board was used in two cases, while the users did not use any other aids besides the VEMOTION® during robot-assisted mobilisation. These aids were used to transfer patients in bed or to transfer them to another therapy device, such as a mobilisation or therapy chair. In three observations, patients were assisted in walking by a forearm walker.
In terms of staffing, two nurses were involved in most mobilisation cases at all survey points (T1 n=20: 57.1%; T2/3 n=44: 68.8%). In 12 cases (34.3%) a nurse mobilised the patients conventionally without the VEMOTION® without further staff support. By comparison, 14 robot-assisted mobilisations (21.9%) were performed by a nurse without further staff support at survey point T2/3. Three or more persons were rather the exception during conventional (n=3, 8.6%) and robot-assisted mobilisation alike (n=6, 9.4%).
Use of kinaesthetic techniques during mobilisation was also examined as well. There were no major differences between the mobilisation methods. Use of kinaesthesia was observed in 6 (17.1%) nurses during conventional mobilisation. 10 (15.6%) nurses used kinaesthetic techniques for mobilisation with the VEMOTION®.
Posture of the mobilising specialist staff
The most frequent variation of each posture of the mobilising person was documented when observing the postures of the mobilising nurse.
Table 7 shows the flexion in the lumbar spine and thoracic spines of the users during survey points T1 and T2/3. Only a marginal difference for flexion in the lumbar spine was found between conventional (n=14, 40.0%) and robot-assisted mobilisation (n=27, 42.2%). Upper body inclination in the thoracic spine to 20-60 degrees was also nearly unchanged in T1 (n=19) with 54.3% and in T2/3 (n=35) with 54.7%. The lowest tilt of the upper body of <20 degrees was observed in 40.6% of robot-assisted mobilisations (n=26) and thus slightly more frequently than during conventional mobilisation (n=12, 34.3%).
(Table 7)
In the observation sheet evasive movements of the upper body were defined as lateral flexion of the upper body, rotation at the waist, their combined movement, or no evasive movement. During 14 conventional mobilisations (40.0%) and 20 mobilisations with the VEMOTION® (31.3%), a combined evasive movement to the side and rotating at the waist was observed most frequently. An evasive movement exclusively rotating at the waist was only evident in six of the conventionally performed mobilisations (17.1%) while it was found in 13 of the robot-assisted mobilisations (20.3%).
The observation of the leg posture was recorded based on the flexion or extension of the knees, the position of the feet in relation to each other, and the angle at which the mobilising specialist was standing in relation to the bed. The nurses stood at an angle of more than 90 degrees to the bed only once during both conventional and robot-assisted mobilisation (T1: 2.9%; T2/3: 1.6%). A position parallel to the bed was documented most frequently in 53 robot-assisted mobilisations (84.1%). This parallel position to the bed was also chosen most frequently by the nurses during conventional mobilisation (n=26) with 74.3%.
In both forms of mobilisation, the knees of the mobilising persons were rather extended than bent, but in T1 (n=26) to a higher proportion (78.8%) than during mobilisation with the VEMOTION® (n=42, 66.7%). The fencer stance was recorded only once at T1 (2.9%), and during 8 robot-assisted mobilisations in T2/3 (12.7%). Most frequently, a parallel stance could be observed among the nurses during mobilisations (T1 n=26, 74.3%; T2/3 n=45: 71.4%).
The most obvious difference in the observation of the postures was recorded in the position of the shoulders. Shoulder elevation was observed significantly less frequently in the users during robot-assisted mobilisation. Figure 4 below shows that users pulled their shoulders upwards in two-thirds of cases during conventional mobilisation (n=24), while users’ shoulders remained in a neutral position in more than 70% of robot-assisted mobilisations (n=45).
Users’ subjective assessments of psychological stress
The subjective assessment of the mobilising specialist staff regarding their psychological stress (PSaR) during mobilisation was performed using a 10-point numerical scale (0=no stress, 10=very severe stress). No relevant deviations between the survey points are evident here (T1 n=35: =3.1 ±3.9/3.09, Md=3.0, SD=1.884; T2/3 n=64: =3.33 ±6.67/3.33, Md=3.0, SD=2.564).
Figure 5 shows the distribution of the subjectively assessed psychological stress (0= no stress, 10= very strong stress) differentiated in accordance with the professional qualification of the nurses with and without further training at the survey points T1 and T2/3.
When differentiating in accordance with the professional qualification of the nurses, the position measures in T2/3 showed marginal differences in robot-assisted mobilisation between the groups of nurses with and without specialist training. The Mann-Whitney-U test shows no statistical significance here (see table 8).
(Table 8)
The correlation between intensive care experience in years and age of the nurses and their subjective assessment of psychological stress and relief (PSaR) was tested using the Kendall-Tau-b correlation coefficient. There were no significant results in T1 during conventional mobilisation. At the point of the survey T2/3, however, a weak negative correlation was found between the nurses’ time of experience in an intensive care unit and their perception of psychological stress. The observed nurses with a longer period of experience in intensive care showed a subjectively lower psychological stress during robot-assisted mobilisation. Table 9 shows the results of the correlation coefficient Kendall Tau b of the variable “Subjective assessment of psychological distress and relief” (PSaR) with age, and intensive care experience (ICU) in years of the mobilising nursing staff (users) at both survey points.
(Table 9)
Furthermore, subjective assessment of the feasibility of the mobilisation was additionally requested using a 7-point numerical scale (0=not feasible at all, 7=very feasible) at the point of the survey T2/3 in order to detect any changes in the psychological stress during robot-assisted mobilisation with the VEMOTION®. Figure 6 reflects the distribution of the subjectively assessed feasibility according to nurses with and without specialist training during robot-assisted mobilisation at survey point T2/3.
When considering all users without differentiation in professional qualification, it was found that the persons who predominantly rated the feasibility of mobilisation with the VEMOTION® as high assessed themselves as less psychologically stressed than persons who felt less able to perform mobilisation with the VEMOTION® (PSaR/feasibility: Kendall Tau b=-0.474, p=<0.001, n=63).
Regarding assessment of the feasibility of robotic mobilisation with the VEMOTION® between the groups of nurses with and without specialist training (PD with FWB: n=40; PD without FWB: n=24), however, the Mann-Whitney-U test showed no significant differences (U=458.5, z=-0.306, p=0.760).