High Adherence to Remote Monitoring Technology in Patients at Risk for Diabetic Foot Ulcer

ABSTRACT

Globally, a lower‐extremity amputation occurs every 20 s as a complication of a diabetic foot ulcer, underscoring the urgent need for effective preventive strategies. Previous studies have shown that temperature‐based foot monitoring can reduce both the incidence and severity of diabetic foot ulcers. However, real‐world adherence data for remote temperature monitoring remain limited, particularly in diverse or resource‐constrained communities. We conducted a pilot implementation study of 20 adults with diabetes and a history of diabetic foot ulcers to assess adherence to a remote foot temperature monitoring mat within the context of receiving podiatric care. Participants are instructed to stand on the mat for 20 s daily, and data are transmitted wirelessly for remote monitoring. Adherence was defined as use of the mat at least four times a week. Participants demonstrated high adherence to the foot monitoring mat, averaging 6 scans per week, with sustained adherence over the 6‐month study period. These findings suggest that high‐risk patients with diabetes can reliably engage with the foot temperature monitoring technology, supporting its potential as a management tool to improve outcomes and reduce the burden of diabetic foot ulcer‐related complications in high‐risk, resource constrained patient populations.

Highlights

• Our study demonstrates that high risk patients with diabetes from diverse and resource‐constrained communities are highly and consistently adherent to the uptake of foot mat temperature monitoring.

• These findings support the use of foot temperature monitoring technology as a management tool for diabetic foot ulcer recurrence in this patient population.

1. Introduction

Approximately, 131 million patients globally suffer from diabetes‐related lower extremity complications [1]. Diabetic foot ulcer (DFU) is a common complication due to the loss of protective sensation in the feet, making them vulnerable to unnoticed trauma due to abnormal foot biomechanics when ambulatory or wearing poorly fitting shoes [2, 3]. Patients with end stage renal disease (ESRD), neuropathy with foot deformity, peripheral vascular disease without prior revascularizations, and a history of prior ulcerations and/or amputations are at higher risks of DFU [4]. DFU is known to be a highly recurrent condition with a three year recurrence rate of DFU of 58% and 50% of the patients requiring endovascular reintervention [5]. Therefore, a healed DFU, like cancer, is more conceptualised by the International Working Group on the Diabetic Foot (IWGDF) as ‘DFU in remission’, which requires a mode of reliable continuous surveillance and proactive prevention.

DFU has multiple complications and it is associated with high rates of hospitalizations and lower extremity amputations (LEA), leading to a loss of mobility, unemployment and a 5‐year mortality rate of 50%–70% [6]. DFUs also lead to higher emotional distress and impose a substantial financial burden on both patients and the healthcare system, contributing to $237 billion in diabetes‐related healthcare costs reported in 2017 [7, 8, 9]. Low income has been associated with higher rates of LEA highlighting the role of economic status and DFU outcomes [9]. Diabetes is prevalent in the Bronx, New York, a region with a high rate of poverty [10]. Among previously ambulatory patients in this cohort, below‐knee or above‐knee amputation—a severe consequence of DFU—occurred at a median age of 55 years, with two cases as young as 36, underscoring the substantial burden of this disease in a younger population [11]. Prevention strategies for DFU that are also robust across all income levels would have a major impact on health care outcomes.

A monitoring strategy to identify patients at risk for a DFU is based on the observation that localised increases in foot temperature exceeding 4°F (2.2°C) reflect subclinical inflammation and often precede ulceration [12]. Although home monitoring with handheld infrared thermometers has been shown to reduce DFU recurrence, this approach is often cumbersome and difficult to perform reliably, particularly on the plantar surface of the foot [13]. In contrast, emerging smart mat temperature‐monitoring systems require patients to stand on the device for approximately 20 s daily and have been shown to reduce DFU incidence in high‐risk individuals, improve clinical outcomes, and decrease health care utilisation [14, 15]. Importantly, this approach may reduce common barriers to care in resource‐limited communities, as monitoring occurs at home and does not require internet or cellular connectivity; data are transmitted automatically and wirelessly to the care team, minimising non‐adherence associated with manual temperature logging [16]. However, real‐world adherence data for remote temperature monitoring (RTM) are very limited, particularly in diverse and resource‐constrained communities.

In this implementation‐oriented study, we assessed 6‐month adherence to an RTM foot mat device among patients with diabetes at high risk for recurrent DFUs receiving routine podiatric care. We also evaluated RTM inflammatory alerts, DFU incidence, and hospitalizations.

2. Materials and Methods

This prospective longitudinal study evaluated adherence to a RTM foot mat among adults at high risk for DFU recurrence. The primary outcome was adherence, defined as completing at least four‐foot scans per week. Secondary outcomes included DFU‐related outpatient visits and hospitalizations, assessed through the electronic health record (EHR).

Twenty participants aged 18–75 years with type 1 or 2 diabetes (T1D or T2D) and a documented history of DFU within the prior 24 months were enrolled. Eligible participants were ambulatory and had clinically confirmed peripheral neuropathy. Exclusion criteria included an active DFU and a history of above‐the‐knee amputation or below‐the‐knee amputation. Participants were recruited from a general referral podiatry clinic in the Bronx within Montefiore Medical Center, where they received routine podiatric care.

Participants who met eligibility criteria and provided informed consent were mailed a SmartMat device from Podimetrics (Podimetrics, 100 Dover St, Somerville, Massachusetts 02144). They received standardised remote training from the Podimetrics nursing team and were instructed to stand on the mat for 20 s each day.

The device automatically records plantar foot temperatures at six contralateral sites and securely transmits data to Podimetrics for analysis and storage. During the initial scans, an individualised baseline foot temperature profile was established for each participant. A temperature elevation exceeding 2.2°C above baseline at any of the six plantar sites on two consecutive measurements was classified as a ‘hot spot’, indicating potential subclinical inflammation.

Detection of a hot spot triggered a telephone based outreach from a study nurse to assess for systemic or foot‐related symptoms (e.g., fever, chills, skin changes). Participants were instructed to initiate an offloading protocol that included reducing standing and ambulation by at least 50%, elevating the affected foot, using appropriate footwear, and continuing daily foot examinations.

If clinical symptoms were present or temperature elevations persisted for more than 2 weeks, participants were advised to seek in person medical evaluation from their health care provider. Clinical escalation was defined as: (1) the need for in‐person evaluation by a healthcare provider or (2) persistent temperature elevation above the threshold despite offloading for more than 2 weeks.

A re‐engagement protocol was implemented whereby the study nurse contacted participants by telephone after identification of four or more consecutive days without mat data transmission. At the end of the 6‐month study period, participants were offered an optional survey developed by the research team on their experience using the mat. Participant adherence and temperature data were also extracted.

Participant characteristics were summarised using descriptive statistics using GraphPad Prism (Graph Prism v10.0). The Kruskal‐Wallis test was used to evaluate adherence to the RTM over the 6‐month period. This study protocol was approved by the Institutional Review Board of Albert Einstein College of Medicine.

3. Results

The study cohort included 20 participants with a mean age of 63 years. Most participants identified as Black (60%) or Hispanic (40%), and 85% were male. The mean HbA1c was 6.9% (Table 1). All participants had T2D and were insured by Medicare or Medicaid.

TABLE 1.

Baseline characteristics of the study cohort.

Variables	Values
Age, (years)	62.7 ± 8.0
Male, no. (%)	17 (85)
Race, no. (%)
Black	12 (60)
White	4 (20)
American Indian/Alaska native	1 (5)
Others	3 (15)
Ethnicity, no. (%)
Hispanic	9 (45)
Non‐Hispanic/not specified	8 (40)
Other	3 (15)
BMI, (kg/m2)	28 ± 1.7
HbA1c, (%)	6.9 ± 0.4
Comorbidities, no. (%)
End stage renal disease	5 (25)
Coronary artery disease	5 (25)
Congestive heart failure	5 (25)
Depression, bipolar, or schizophrenia	2 (10)
Stroke/TIA	3 (15)
Peripheral neuropathy	19 (95)
Peripheral arterial disease	13 (65)
Hypertension	16 (80)
Hyperlipidemia	11 (55)
History of smoking, no. (%)	7 (35)
DFU within last 2 years, no. (%)	15 (75)
History of DFU related amputations, no. (%)	17 (85)
Uses mobility aid, no. (%)	12 (60)
Uses therapeutic shoes, no. (%)	7 (35)

Note: Continuous variables presented as mean ± SD, categorical variables presented as percentages.

Abbreviations: BMI, body mass index; HbA1c, haemoglobin A1c.

Nineteen of twenty participants completed the study; one participant was lost to follow up due to a non‐study related death. All the participants who completed the program met the predefined adherence criteria. In addition, there was no significant change in the median number of weekly foot scans across the six‐month period (Figure 1—p‐value = 0.12).

FIGURE 1.

Median number of weekly foot scans by month of study. Number of weekly scans per participant across the 6‐month study period. No significant change over time (Kruskal Wallis test, p‐value = 0.12). Medians and interquartile ranges shown.

The mat detected 14 episodes of inflammation during the 6‐month period among 12 participants. Six participants experienced a single clinical escalation requiring medical intervention for a DFU‐related complication. These events included five cases of DFU and one case of painful pre‐ulcerative hyperkeratosis associated with a plantar callus. Of the five DFU cases, four occurred on the plantar surface of the foot, while one was a chronic non‐pressure‐related ankle ulcer.

Among the four plantar ulcers, inflammatory changes were detected by the mat prior to ulcer development in two cases, prompting outreach and offloading recommendations. The remaining two plantar ulcers were not preceded by detectable inflammation. One involved a distal superficial wound of the left fourth toe, which, due to its distal and small size, is difficult to visualise with RTM technology. The other developed within the first weeks after enrollment and may have reflected subclinical inflammation that occurred prior to study onboarding.

Six hospitalizations occurred during the study period. Only one hospitalisation was attributable to a DFU in which the participant developed osteomyelitis. No participants had an LEA during the study period.

Thirteen study participants required re‐engagement outreach during the study period, with many requiring multiple contacts, for a total of 40 calls. All participants who developed a plantar DFU received at least one re‐engagement call, and four of the five participants who developed a DFU had two or more re‐engagement calls. However, repeated re‐engagement calls (≥ 2 calls) were not significantly associated with development of a plantar ulcer (χ ² test, p‐value = 0.72).

Of the 19 participants who completed the 6‐month trial period, 11 completed the optional survey. All participants rated the mat as ‘very easy’ or ‘easy’ to use and reported being ‘very satisfied’ or ‘satisfied’ with the service provided. When asked whether they would continue the program if the option was available, seven participants said yes, two responded, ‘I'm not sure’, and two participants said ‘no’.

4. Discussion

Our pilot study demonstrates that high‐risk patients with a history of DFU can adhere to foot monitoring using smart mats and maintain adherence over time. Participants experienced frequent episodes of foot inflammation, with 63% demonstrating clinically significant temperature elevations. Overall, participants reported high satisfaction with the use of the RTM device.

There were 14 episodes of persistent plantar temperature elevation, reflecting a high frequency of subclinical inflammation. Two participants with documented hotspots subsequently developed DFUs, whereas the remaining 12 did not, suggesting a potential protective signal consistent with prior studies, though this study was not powered to test efficacy [14, 15, 17]. Progression to DFU in participants without detected hotspots may reflect limitations in temperature monitoring, including regions of the foot that are difficult to reliably assess, or the presence of subclinical inflammation prior to enrollment. Conversely, in participants who developed DFUs despite detected hotspots, adherence to offloading recommendations and the quality of prescribed footwear may have influenced outcomes and warrant further investigation.

Overall, participant satisfaction with the device was high. Individuals with diabetes often experience substantial distress related to the risk of complications, including major LEA and infection, as well as the ongoing demands of self‐care [18, 19]. It is possible that the use of RTM provides reassurance, thereby reducing diabetes‐related stress and further encourages self‐care behaviours. Further research is needed to evaluate the impact of RTM on diabetes‐related distress and self‐management outcomes.

Our pilot findings complement the existing RTM literature, although we acknowledge limitations related to sample size and event rates. In the Rothenberg Veteran Affairs Medical Center study, 70% of participants demonstrated full engagement after a year of monitoring [17]. Similarly, in the Frykberg multicenter study, 132 participants with diabetes and prior DFU were followed for 34 weeks and 86% used the mat at least 3 days per week and 88% reported that the mat was easy to use [20]. In this context, the adherence observed in our 20‐patient pilot corroborates and further refines the growing body of evidence demonstrating that patients at high risk for DFU can sustain regular at‐home use of RTM technologies over time.

Our cohort differed from prior RTM mat studies in several important respects, including inclusion of a historically underserved population, a different geographic setting, and management under a single prescribing podiatrist. Participants were drawn from a regional referral center for diabetic foot complications, receiving referrals from vascular surgery, endocrinology, emergency departments, primary care, podiatry, orthopaedics, infectious disease, and home care nursing teams across multiple campuses of a large urban healthcare system. As a result, the cohort reflects the racial, ethnic, and socioeconomic diversity of the Bronx, consistent with the borough's demographics.

The high adherence observed in this study supports the potential generalizability of RTM adoption in a diverse, low‐income population, extending findings from prior studies. While these distinctions may influence adherence behaviours and limit direct comparability with earlier studies from controlled settings, the persistence of adherence to RTM suggested that this technology is well tolerated and can be readily integrated into patients' daily routines. By examining adherence within this real‐world care environment and operational framework, our findings help narrow the gap between controlled RTM trials and everyday clinical practise.

A recent retrospective study by Littman et al. did not find that RTM was associated with reduction in LEA or all cause hospitalisation [21]. Although no participants in our study underwent LEA, the relationship between RTM‐detected potential ulceration and subsequent amputation outcomes remains unclear. It is also important to note that patients with a history of DFU are conceptualised by the IWGDF as being in remission rather than cured, necessitating longitudinal risk‐management within a coordinated care framework.

Accordingly, adherence to RTM technology represents only one component within the multifactorial processes that influence LEA rates. Effective DFU prevention and management are also highly dependent on the broader healthcare infrastructure supporting the patient. While both our study and prior studies demonstrate high adherence to RTM, our study was conducted within a single provider framework whereas Frykberg included multiple foot care centers, which may contribute to differences in downstream outcomes such as LEA and mortality rates [20].

Future studies should further investigate additional factors that may influence the effectiveness of RTM, including adherence to offloading, socioeconomic barriers to care, the presence of a reinforcement or reminder system, variability in healthcare infrastructure, and heterogeneity in DFU treatment and follow‐up. These factors may collectively influence RTM effectiveness, DFU development, and LEA incidence.

The smart mat technology has several limitations. Ulcers may develop in non‐plantar or poorly monitored regions of the foot, and appropriate offloading may not be feasible for all patients. In addition, key contributors to DFU, such as abnormal foot biomechanics, motor neuropathy, and prior digital or midfoot amputations that increase plantar pressure, are not captured by RTM [3].

Our study also has limitations. The absence of a control group precludes assessment of the mat's efficacy in predicting DFU, limiting causal inference. Furthermore, the cohort demonstrated excellent baseline diabetes self‐care, with a low mean baseline HbA1c and high use of therapeutic footwear, which may limit generalizability of adherence to broader high‐risk populations. Adherence to offloading recommendations was not systematically assessed and may be an important unmeasured factor influencing ulcer development despite recommended offloading.

Despite these limitations, our findings demonstrate that high‐risk patients with diabetes in an urban setting can be highly engaged with and adherent to an RTM‐based DFU prevention program. This suggests that consistent smart mat use is feasible in real‐world, historically underserved populations and may achieve levels of engagement comparable to benefits similar to those observed in controlled studies.

Funding

This work was supported by New York Regional Center for Diabetes Translational Research (342843) and American Diabetes Association (CDTR #5).

Ethics Statement

The protocol was approved by the Albert Einstein College of Medicine Institutional Review Committee, IRB# 2023‐14915.

Consent

Informed consent was obtained from all participants.

Conflicts of Interest

Gary Rothenberg: Director, Medical Affairs, Podimetrics. The other authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.