Footwear and insole design features for offloading the diabetic at risk foot—A systematic review and meta‐analyses

Abstract The aim of this systematic review was to identify the best footwear and insole design features for offloading the plantar surface of the foot to prevent foot ulceration in people with diabetic peripheral neuropathy. We searched multiple databases for published and unpublished studies reporting offloading footwear and insoles for people with diabetic neuropathy and nonulcerated feet. Primary outcome was foot ulcer incidence; other outcome measures considered were any standardized kinetic or kinematic measure indicating loading or offloading the plantar foot. Fifty‐four studies, including randomized controlled studies, cohort studies, case‐series, and a case‐controlled and cross‐sectional study were included. Three meta‐analyses were conducted and random‐effects modelling found peak plantar pressure reduction of arch profile (37 kPa (MD, −37.5; 95% CI, −72.29 to −3.61; P < .03), metatarsal addition (35.96 kPa (MD, −35.96; 95% CI, −57.33 to −14.60; P < .001) and pressure informed design 75.4 kPa (MD, −75.4 kPa; 95% CI, −127.4 to −23.44 kPa; P < .004).The remaining data were presented in a narrative form due to heterogeneity. This review highlights the difficulty in differentiating the effect of different insole and footwear features in offloading the neuropathic diabetic foot. However, arch profiles, metatarsal additions and apertures are effective in reducing plantar pressure. The use of pressure analysis to enhance the effectiveness of the design of footwear and insoles, particularly through modification, is recommended.

foot ulceration may be prevented. 10 Elevated dynamic plantar pressures during locomotion contribute to the development of plantar diabetic foot ulcers when in the presence of neuropathy. 11,12 Guidelines recommended that people with diabetes wear appropriate 'diabetic footwear' designed to reduce repetitive stresses at all times. 13 Systematic reviews have demonstrated the effectiveness of footwear and insoles in offloading the plantar load under the foot and preventing ulceration. [14][15][16][17][18] However, these have not identified the best insole design or feature and footwear specification or modification for use when reducing plantar load for foot ulcer prevention in people with diabetes and neuropathy.
Therefore, the purpose of this systematic literature review was to identify the best footwear and insole design features for offloading the plantar surface of the foot to prevent foot ulceration in people with diabetes. It is anticipated that this information will inform a standardized protocol for the clinical design of therapeutic insoles and footwear to offload the foot and reduce ulcer risk in people with diabetes and neuropathy.
More specifically, the objectives are to identify the key design features with regard to the following: • profile/shape of the insole, shoe upper and shoe outsole • material type and properties of the insole and shoe outsole • modifications made to the insole and shoe outsole • fabrication techniques used for the insole and shoe

| ME THODS
This systematic review was performed and reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Guidance. 19 The systematic review was prospectively registered on the PROSPERO database for systematic reviews (CRD42017072816).
The population of interest was adults over 18 years of age with type 1 or 2 diabetes mellitus and peripheral neuropathy. The primary outcome was foot ulcer incidence; other outcome measures considered were any standardized kinetic or kinematic measure indicating loading or offloading the plantar foot (such as plantar pressure, pressure-time integral, total contact area, dynamic measures of centre of pressure trajectory or velocity) and any standardized clinical measure indicating loading/offloading of the plantar foot (such as callus/lesion reduction).
Side effects/adverse events as a result of the design features were additional outcomes of interest. We excluded studies on people with active ulceration, major amputation of the foot or Charcot arthropathy because we considered that the unique pathomechanics and gross deformity associated with the severity of these conditions would unduly influence the design features of the footwear and insoles.
This review included both experimental and epidemiological study designs including randomized controlled trials, non-randomized controlled trials, quasi-experimental, before and after studies, prospective and retrospective cohort studies and analytical cross-sectional studies.
Studies were included if they made one of the following comparisons: footwear and/or insole design feature compared with another therapeutic footwear and/or insole design feature; footwear and/or insole design feature compared with no intervention. Qualitative studies, case reports and systematic reviews were excluded. specialist. An example of the search from one of the databases is provided in Appendix S1. Title and abstract of all papers retrieved by the literature search were screened independently by two researchers (RC and JP) to determine whether the paper met the inclusion criteria with disagreements resolved by discussion. Full-text articles were then retrieved and further screened by two researchers (RC and JP) independently for inclusion in the review. In addition, a hand search was undertaken using the references from journal articles.

| RE SULTS
The initial electronic search generated 7384 articles of which 2094 were duplicates (Figure 1). In the screening phase, 4750 were excluded based on their title and a further 466 excluded on title and abstract leaving 74 articles for full text assessment. We excluded 28 of these articles based on irrelevant study population (n = 12), irrelevant study design (n = 4), irrelevant outcome/ intervention (n = 12) leaving 46  included in the final review. As the initial search was undertaken in July 2016, updated searches were performed in December 2017 yielding 6918 articles, from which an additional three studies [66][67][68] were included and November 2019 yielding 7821 articles from which a further five studies 69-73 were included.

| Data extraction
Data extraction of included studies was conducted using JBI Meta-Analysis of Statistics: Assessment and Review Instrument (JBI-MAStARI). 74 In this phase, the general and contextual data were extracted in relation to the population, study design, interventions (features, design, modifications and materials of footwear and insoles) and outcomes. In addition, relevant information was extracted in the results section. Data extraction was carried out by (RC) and checked by the second reviewer (JP).

| Data analysis and synthesis
In this review, we summarized study findings quantitatively and pooled study effects in a meta-analysis when appropriate using JBI MAStARI. 74 Meta-analysis was performed using random-effects models for continuous variables, calculating mean differences using the inverse variance method. Meta-analysis was based on changes from baseline for peak pressure when the mean and SD were reported where any footwear or insole design feature, modification and material or method could be distinguished. Means and SD's of data were required to be included in the meta-analysis; we contacted four corresponding authors to request this data when not included in the article; two authors did not respond, and one no longer had access to the data.

| Assessment of study quality
Two reviewers (RC and JP) independently assessed the methodological quality of the studies using the relevant JBI critical appraisal tools. 76 Disagreements were resolved through consensus meeting.
A study was considered low risk of bias if all criteria was included.
Summaries of the appraisal of study quality are included in Appendix S2. All studies had some form of bias with standards of reporting variable across studies and by study design. From the quality assessment of the randomized controlled trials (RCT's, all of the RCT studies had some form of bias (mean percentage of 'yes' scores = 65% ±SD 29%). All RCT studies reported inclusion criteria of participants, p values and participants lost to follow-up. The most frequent omissions related to the blinding of the assessor and participants, concealing of treatment allocation and outcomes measurement.
Follow-up time periods ranged from no follow-up to five years.
Only 10% (n = 5) repeated measure studies 21,24,36,43,60 measured the direct effect of an arch profile on mean peak pressure. According to the heterogeneity test, high heterogeneity existed (I 2 = 81%, χ 2 = 13.6, τ 2 = 1160, P = .009). Therefore, random-effects modelling was applied to consolidate the effect value. Figure 2 shows that that out of 119 participants, the addition of an arch profile reduced peak Four studies reported the effect of a rocker profile. One study reported that in 71%-81% of participants tested an optimum peak pressure target value of under 200 kPa could be achieved with a combination of apex position at 52% of shoe length and rocker angle of 20°. 67 Another study reported no interaction effect when altering apex angle, apex position and rocker angle compared with the control shoe. 30 A third study reported decreases in peak pressures and pressure-time integrals in the posterior and anterior, central lateral and central medial forefoot with a standardized rocker shoe with apex position (83 mm on medial and 87 mm on lateral from front of shoe), angle thickness (24 mm maximum thickness at rocker with 11 mm rocker height at front end) compared to shoe without rocker. 40 A fourth study reported ulcer reoccurrence to be 64% with a semi-rigid rocker sole compared to 23% with a rigid rocker sole. 70 There was an inability to distinguish the effect of the rocker profile feature from other features of the footwear and insole for those remaining studies. 26
Therefore, random-effects modelling was applied to consolidate the effect value. Figure 3 shows that out of 70 participants, the use of a metatarsal addition in an insole reduced mean peak pressure by a fur- Thirteen per cent (n = 7) of studies 27,31,33,36,42,73,77 used 'other' modifications. One study reported a 71% reduction on ulcer incidence when using 'intelligent' insoles with pressure detecting sensors compared to the control group. 77 One study reported a 9 kPa reduction in mean peak pressure when adding a custom-made five degree full length varus and valgus cork posts to the base of the insole for 20 participants with diabetic peripheral neuropathy and nondeformed feet. 36 The remaining studies did not report the effect of these modifications. One study reported balancing the ¾ length orthotic with the use of dental acrylic posts at the rearfoot 31 and another study used extra-density padding at the heel, forefoot and covering the toes as a modification. 33 Another study reported the use of wedge or medial skive on two occasions, prescribed at the discretion of an orthotist, but no rationale for use provided. 73 One study reported including elastic binders and two nonstick sheets placed between the upper and lower pad of the insole as part of their shear resistant insole, 42 and one study used substantial heel cups in the design of their insole, although no specification was disclosed. 27  26,43,48,64 reported the direct effect of using plantar-based pressure analysis as a fabrication technique to inform the design and modification of the insole and shoe in reducing mean peak pressure.

| D ISCUSS I ON
The aim of this review was to identify the best footwear and insoles design feature for offloading the plantar surface of the foot to prevent foot ulceration in people with diabetes. More specifically, the objectives were to identify the key design features of footwear and insoles with regard to profile and shape, material type and properties, modifications and fabrication techniques.

| Modifications
The

| Fabrication techniques used for the insole and shoe
Two different fabrication techniques for insoles and footwear were identified in this review; casting and kinetic informed.
Casting is traditionally used to capture the geometric shape of the patient's' foot to 'customize' the insole. Only one study examined the role of three types of casting technique in reducing peak pressure. 60 The authors reported an insole formed from a semi-weightbearing foot shape offered the greatest peak pressure reduction compared with full-weight-bearing and non-weight-bearing foot shapes, but was not statistically significant. The remaining studies using a casting approach were not able to report any difference in reducing pressure using this fabrication method. This method of fabrication is believed to create an arch profile, which has been demonstrated as altering pressures in the plantar foot as reported by four studies. 21

| Material type and properties of the insole and shoe outsole
Material choice is an important feature of any insole or footwear design. The material used, dependent on its mechanical and physical properties, will influence the insole or footwear's ability to redistribute or dampen forces effectively. This review found no consistency with individual materials used or thickness in the construction of footwear or insole. Only one study directly assessed the effect of material hardness in reducing peak plantar pressures. 38 Sixty-seven per cent of remaining studies used either dual or multi-density material constructions of footwear and insoles. Closed cell foam materials were most frequently sited at the interface between foot and insole and footwear as a top cover; denser materials constituted the base of the insole, EVA appearing the most popular material of choice for the base. A less popular material type was thermoplastics, potentially because these materials were traditionally used for functional devices aimed towards changing gait function and not reducing pressure. Combining materials of different properties is suggested as incorporating the desired properties from each material to best serve reduction in foot ulceration risk. [84][85][86] However, the literature does not provide a sufficiently robust evidence base to inform the selection approach regarding material combination or thickness for the best offloading. Therefore, selection of materials is often influenced by the availability of materials locally and anecdotal evidence, rather than patient-specific characteristics and effectiveness of offloading.

| LI M ITATI O N S
The primary limitation of this review is the heterogeneity of study

ACK N OWLED G EM ENTS
I am grateful for the assistance of the University of Plymouth librarians for their assistance with the search strategy.

E TH I C S A PPROVA L
This review manuscript summarizes and informs of already published studies and thus does not require ethical approval.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.