In this pilot study, we set out to measure the distribution and localization of support surface interface pressures in neonates in a PICU. This included mechanical testing of pressure peaks to compare the state-of-the-art foam mattresses used in the clinics and a new prototype air mattress. This preliminary work was performed in order to develop a prototype mattress to reduce PIs in this patient population.
Setting and patients: The measurements were conducted in the 27-bed level III PICU of a children's university hospital in Switzerland between November and December 2020. During this period, all regulations with regard to Covid-19 safety measures were followed. During the pressure measurements, every neonate was cared for routinely by the local PICU staff; additionally, experienced PICU-trained nurse supervised the intervention to ensure the comfort of the neonates. The nurse was further in charge of assessing symptoms of discomfort in the neonates, interrupting the measurements if necessary. All measurements were approved by the responsible physician and supervised by medical specialists, while maneuvers with the neonates were carried out exclusively by the medical personnel. Parents of the neonates were invited to accompany their child during the procedure and could interrupt the procedure at any stage. All measurements were conducted in the evening shift between 4 and 10 pm. The measurements were conducted on five neonates who had been admitted to the PICU due to organ failure during their first month of life.
Ethical approval: The pilot study was assessed by the ethics committee of the hospital as well as all involved parties within the PREPICare consortium (BASEC 2020-01183; ETH: EK 2020-N-153). Informed consent was provided by the parents.
Interventions and materials: PIs are defined by the European Pressure Ulcer Advisory Panel (EPUAP), the National Pressure Injury Advisory Panel (NPIAP, formerly National Pressure Ulcer Advisory Panel), and the Pan Pacific Pressure Injury Alliance (PPPIA) as
…localized damage to the skin and underlying soft tissue usually over a bony prominence or related to medical or other devices. The injury can be present on intact skin or as an open ulcer and may be painful. The injury occurs as a result of intense and/or prolonged pressure or pressure in combination with shear. The tolerance of soft tissue for pressure and shear may be affected by microclimate, nutrition, perfusion, comorbidities, and condition of the soft tissue. [7,8]
This general definition defines all PI types and encompasses various causal factors.
We measured pressure distribution for a standard foam mattress (Babytherm 8004, Dräger, Lübeck, Germany) as well as for the prototype of a laser-welded air-filled mattress developed by Empa, Swiss Laboratories for Materials Science and Technology, St. Gallen, Switzerland.
Air mattress prototype development: laser welding (as described previously by Fromme et al. [28]) was used to form air-filled structures based on thermoplastic polymers. Polyvinylchloride (PVC)/ polyurethane (PUR) angle connectors (Carmo A/S, Espergærde, Denmark) were welded into each chamber to allow connection to tubes for inflation of the created 3D structures. An iterative approach to prototyping and testing in the lab was used to optimize the structure for testing in the clinical setting. The structure was formed of three unconnected air chambers, also referred to as zones. Different types of membranes and laminates were explored, with the final version being made from a Sympatex polyester laminate (Sympatex 150 STX Orinoco, Sympatex Technologies GmbH, Unterföhring, Germany) (Figure 1). An electronic unit consisting of a microprocessor, pump, several valves, and multiple pressure sensors allowed for exact pressure regulation in the different zones.
Measurement set-up: The air mattress was placed on top of a standard foam mattress for the measurements. Numeric analysis of the pressure peaks was carried out, and representative MATLAB color maps were produced for visual comparison of the deflated and inflated states.
A flexible pressure-sensing mat (PSM; Xsensor LX100:40.40.02, XSENSOR Technology Corporation, Calgary, Canada) with a matrix of 40 x 40 sensing units and spanning an area of 508mm x 508mm was used for the pressure measurements. The measurement range was 5–200 mmHg, with an accuracy of 10% in this range; values below 5 mmHg were excluded automatically by the electronics and software of the PSM.
Pressure distributions for the neonates on the mattress system were measured on-site. Before the start of the experiment, we set the baseline by lying the neonates on a standard foam mattress (Babytherm 8004) with all other positioners (e.g., towels) removed. A thin medical sheet (curea medical GmbH, Steinfurt, Germany) was allowed for the capture of bodily fluids and protection of both the baby (in diapers) and the sensor mat (PSM). The measured interface pressures were constantly monitored by computer. The neonates were in a supine position for all measurements. The measurements were split into four sections, and four minutes of data were collected for each section.
The measurements taken for the first neonate also served as verification of the setup, as initial testing of the air mattress could only be carried out using a baby doll. As a result of the first measurement with a neonate, a small adaptation was made to the air mattress: a small circular seam at the center of the head zone was welded into the mattress to decrease central bloating in this chamber and stabilize the baby's head (Figure 1).
Sequence of measurements to assess pressure distributions (four minutes of data collected per stage) :
- Baseline (sequence 1): the nursing staff positioned the neonate on the standard foam mattress equipped with the PSM and the medical sheet. The PSM was checked for potential folds or overstretched areas.
- Air mattress, deflated (sequence 2): the neonate was lifted, together with the medical sheet and PSM, and the air mattress was positioned in between the PSM and the standard foam mattress in a deflated state.
- Inflation of air mattress (sequence 3): the air mattress was inflated gradually, zone by zone, in multiple cycles, until the neonate was markedly raised. This inflation took up to 10 minutes. With the neonate in a supine position, body parts in the three different zones remained horizontally aligned.
- Air mattress, steady state (sequence 4): after the inflation state was reached, another four-minute steady-state measurement was taken.
Data and statistical analysis
The pressure data captured by the PSM was registered with a sampling frequency of 5 Hz, and MATLAB software was used for data processing and analysis. The 40 x 40 value matrices corresponding to the measurement frames were saved in 3D structures for analysis (40 x 40 x time). Illustration of the pressure data by colormaps allowed easy recognition of zones with elevated pressure. The colormaps corresponded to the data matrices, whereby the pressure value of every sensel was represented by a colored square according to pressure level. Histograms were used to show the differences in pressure value frequencies for the entire body and the occiput in particular. Boxplots were used to compare the baseline and steady states for all neonates; the first box plot included all values ≥ 5 mmHg, while a second box plot included only the highest 10% of interface pressure values (representing the potentially problematic peak pressure values). This percentage was arbitrarily defined since there is no reference for “high” interface pressures for neonates, as mentioned previously [18].
Outliers in the box plot are defined as values 1.5 times higher or lower than the maximum or minimum values in the interquartile range, respectively. Boxes whose notches (median line) did not overlap had dissimilar medians at a 5% significance level. For a non-visual statistical analysis, the mean values of pressures per neonate and sequence were calculated and compared. The Wilcoxon signed rank-sum-test was used to test the significance of the effect at a 5% significance level. This analysis was performed for both data sets, i.e., the full set and the set including the highest 10% of values only. Percentage differences between the baseline and steady-state means for each neonate were calculated to show the size of the effect.