Eight (8) FE model variants were developed to investigate the effects of the design (particularly, the stiffness anisotropy) of prophylactic dressings on the mechanical states of the soft tissues at the sacral region in a supine weight-bearing position (see Table 1). The following conditions were simulated: 1) without a dressing, 2) with an isotropic multilayer compliant dressing, 3) with an anisotropic multilayer compliant dressing, and 4) with a completely stiff isotropic dressing. These mechanical states in soft tissues were simulated in pure compression loading of the buttocks (ie, simulating a horizontal supine bed rest) and in combined compression and shear loads applied to the buttocks (ie, 45˚ Fowler position).
Geometry. In order to develop a 3D, anatomically realistic geometrical model of the buttocks of a supine subject, 76 T1-weighted axial MRI slices were used. A 28-year-old healthy woman was scanned in a supine position, fully weight-bearing, on a designated rigid platform. Imaging was performed in a 1.5 Tesla MR system (MAGNETOM Aera, SIEMENS AG, Munich, Germany) utilizing T1-weighted images (TR/TE=550/10, field of view 42 mm × 420 mm, slice thickness 3 mm), at the Assaf Harofeh Hospital (greater Tel Aviv area), Israel. The MRI study was approved by the Institutional Review Board (Helsinki Committee) of Assaf Harofeh Hospital (Approval no. 190/14). The above MRI scan captured the entire region of the pelvis from the iliac crest to the shaft of the femurs. The image set was imported from the MRI to the ScanIP 3D image module of Simpleware® (Exeter, UK), where semi-automatic segmentation was performed in order to distinguish between the pelvic bones and soft tissue regions16 (see Figure 1A).
Next, 3 of the 5 layers of the isotropic and anisotropic multilayer dressings were applied in the modeling — namely, the polyurethane foam (PUR), the nonwoven (NW), and the airlaid (AL) layers, using the 3D image module. The innermost Safetac® layer (Mölnlycke Health Care) then was added as a tied interface between the skin and the polyurethane foam, preventing these layers from penetrating or sliding across each other in the modeling, and the backing film layer also was represented to address frictional sliding with the elastic (foam) support (as reported in previous papers from the authors13,14).
The present modeling challenge of representing the modes of action of the sacral prophylactic dressing involved allocating greater computational power than demonstrated in the authors’ previous work13,14 involving heel dressings. This is primarily due to the complexity of this 3D FE problem that includes elements with dimensions that vary from fractions of mm for the dressing components and up to tens of cm for the bone and soft tissue structures of the 3D buttocks. Accordingly, several measures were taken to simplify this large deformation problem to the extent that adequate numerical solutions could be obtained, despite the considerable challenge regarding the multiscales as explained previously. First, for these modeling procedures, skin, muscle and fat components were considered together and grouped as “soft tissue” structures. Second, the model volume for the FE analyses was decreased to include the dressing, the sacrum, and the surrounding soft tissues contained in the 3D block shown in Figure 1B. Adequate margins of soft tissue structures were intentionally kept around the dressing to avoid any boundary or edge effects (see Figure 1B).
Next, a flat standard foam mattress was added under the modeled buttocks (and under the dressing, in cases where a dressing was applied). Final FE meshing also was performed in the 3D imaging module using 139 964 to 212 585 linear tetrahedral elements describing the bones and soft tissues as well as 1 636 013 linear tetrahedral elements describing the 3 physical layers of the multilayer dressings. Hence, the FE analyses, which are described here, were conducted using meshes that contained nearly 2 million elements, which was essential given the multiscale challenge, and specifically, for adequate numerical transition between the microscale of the layered structure of the dressing and the macroscale of the buttocks tissues.
Mechanical properties of the tissues and dressing. The constitutive laws and mechanical properties of all tissues were adopted from the literature based on empirical data.13,14 Specifically, the pelvic bone and femurs were assumed to be linear-elastic isotropic materials with elastic moduli of 7 GPa and Poisson’s ratios of 0.3. All soft tissues were considered together as 1 effective material as previously noted17 and were assumed to be nearly incompressible nonlinear isotropic materials, with their large deformation behavior described by an uncoupled Neo-Hookean constitutive model.
The material constants reported by Oomens et al17 were used to represent the effective soft tissue stiffness, assuming that skin contributes 60% to the effective stiffness and the other 40% are attributed to fat. The PUR, NW, and AL layers of the isotropic multilayer dressing were considered isotropic linear-elastic materials with elastic moduli of 24 kPa, 150 kPa, and 30.6 kPa, respectively, based on measurements previously performed in the authors’ laboratory and recently reported.13,14 The Poisson’s ratio assigned to these dressing layers was 0.258 based on published experiments.12 In cases where the completely stiff isotropic dressing was used (variants 4 and 8), the PUR, NW, and AL layers of the stiff dressing were considered isotropic linear-elastic materials with elastic modulus of 1 MPa and a Poisson’s ratio of 0.258. The mattress was considered isotropic linear-elastic as well, with an elastic modulus of 50 kPa and Poisson’s ratio of 0.3, again based on literature.13,14,18
The anisotropic multilayer compliant dressing design comprises anisotropy — directional stiffness properties that constitute a stiffer longitudinal behavior in the direction of the spine versus more compliant “wings” that facilitate lateral stretching of the dressing. To capture this anisotropy feature, the stiffness properties of the PUR, NW, and AL layers of model variants 3 and 7 of the isotropic dressing were increased by 45% only in the axial (Z) direction to replicate the longitudinal stiffness characteristic of the anisotropic dressing based on measurements preformed in the authors’ laboratory to quantify this anisotropy (see Table 1).
Body loads applied to the buttocks model, shear, and friction conditions. Boundary conditions were chosen to simulate the descent of the weight-bearing pelvic bones during supine lying or a 45˚ Fowler position. The response of soft tissues to this descent was tested without and with each of 3 test dressings of the same shape. In all simulation cases, dressings were attached to exactly the same sacral region, ideally aligned, and symmetrically placed according to manufacturer’s guidelines (as these dressings would have been in a real-world scenario), as detailed in Table 1.
In terms of other relevant constraints, the bottom surface of the mattress was fixed for all translations and rotations. Tied interfaces were defined between the bones and soft tissues as well as between the soft tissues and the PUR layer of the dressing to account for the full adherence properties of the Safetac layer of the dressings. Frictional sliding was defined between the AL layer of the dressings and the mattress, with the coefficient of friction set as 0.35 to simulate the low-friction effect provided by the backing film layer of the dressings.12 In model variants 1 and 5 (ie, simulations of the weight-bearing buttocks without a dressing), the coefficient of friction between the soft tissues and the mattress was set to be higher (0.4) because of the absence of the backing film.13
To simulate loading conditions, downward displacements in the range of 5.3 mm to 6.45 mm in all model variants were applied on the top surface of the reduced model volume (marked in Figure 1B) until the total reaction force acting back from the mattress reached 40 Newtons (roughly 7% of the total bodyweight of the subject), which were assumed to be transferred through this reduced model volume for the purpose of comparison across model variants. In model variants 5, 6, 7, and 8, the same extent of displacement also was applied in the axial (Z) direction, accounting for the shearing forces that may act due to sliding down in the bed (eg, when seated in bed in a 45˚ Fowler position) or due to some spontaneous movements or repositioning of the patient in the bed. The FE simulations all were created using the PreView module of FEBio (version 1.18), analyzed using the Pardiso linear solver of FEBio (version 2.3.1), and post-processed using PostView of FEBio (version 1.919) (University of Utah, Salt Lake City, UT).
Biomechanical outcome measures. Volumetric exposures of the soft tissues adjacent to the sacral bone to sustained deformations were examined and quantified in terms of the strain energy density (SED) in these soft tissues within the reduced model volume (see Figure 1B). Briefly, SED is a scalar measure in units of mechanical stress (eg, kPa) that describes the spatial dispersion of the elastic energy that is stored in an object that undergoes deformation. It is a factor of the stiffness of the material and of the mechanical strains and stresses that develop in every point within the deformed object.
Data analysis. The SED data were pooled from the soft tissues for all the elements in a 67 mm x 55 mm x 20 mm soft tissue cube located immediately under the sacrum, which had been defined as the volume of interest (VOI) for the purpose of SED data comparisons across the model variants (see Table 1), as depicted in Figure 1D. Converging time steps were chosen for data collection, so the resulting reaction forces between the buttocks and the support were within less than a 2.4% difference from the aforementioned target reaction force. The SED in the VOI were analyzed across the model variants to determine whether additional biomechanical efficacy is present in the anisotropic multilayer dressing design in terms of alleviation of tissue loads with respect to a no-dressing situation, to an isotropic multilayer dressing case, or to a completely stiff dressing. These simulations were repeated in either pure compression or compression combined with shear loads (see Figure 1C) and compared quantitatively by calculating the volumetric exposures to SED in the soft tissues in the VOI per each simulation case (see Table 1). The details of the method of FE analysis are explained in the authors’ previous publication14 that includes explanations with regard to calculation and data processing techniques.