A custom accelerometer-based algorithm was developed and validated to objectively quantify removable cast walker and contralateral shoe-lift adherence during walking and standing, achieving >99% accuracy in lab testing and excellent agreement with diary-based adherence (ICC >0.96) during 24-hour community monitoring.
Key Findings
Results
The custom algorithm achieved greater than 99% accuracy in detecting steps and standing while participants used the removable cast walker and shoe-lift during in-laboratory testing.
Thirty healthy adults served as participants for algorithm development and validation
Accelerometers were placed on the removable cast walker (RCW), the contralateral shoe-lift, and unilaterally on participants' thighs
In-lab activities were logged by investigators to serve as ground truth for accuracy comparisons
Accuracy was determined by comparing algorithm-derived adherence metrics with investigator log entries
Results
Algorithm-derived adherence showed excellent agreement with diary-based adherence during 24-hour community monitoring.
Intraclass correlation coefficient (ICC) was greater than 0.96 for agreement between algorithm and diary-based adherence
Agreement was statistically significant at p < 0.01
Following in-lab testing, monitoring continued for 24 hours with participants logging footwear use in a diary
ICC > 0.96 is characterized as 'excellent agreement' by the authors
Results
Error rates (misclassifications) for the algorithm were less than 3% during community monitoring.
Misclassification rate of <3% was reported for the 24-hour community monitoring phase
The algorithm classified RCW and shoe-lift adherence based on sample-to-sample variance thresholds
Walking and standing episodes were quantified via the thigh-worn monitor
The low error rate supports the algorithm's reliability for real-world offloading behavior monitoring
Background
The study addressed methodological limitations of prior objective adherence measures by pairing offloading device tracking with contralateral therapeutic footwear (shoe-lift) monitoring during both walking and standing.
The novel method tracks both the removable cast walker and the contralateral shoe-lift simultaneously
The algorithm quantifies adherence specifically during walking and standing episodes, not just total wear time
Thirty healthy adults were used as the study population for device and algorithm development
Conclusions
The validated algorithm is proposed for use in future clinical trials and potentially in clinical practice for monitoring real-world offloading behavior in diabetic foot ulcer patients.
Authors state the method 'will allow for enhanced assessments of adherence in future trials'
The method 'may eventually be incorporated into clinical practice for patient monitoring'
Strong agreement with both lab observations and user diaries supports its use for monitoring real-world offloading behavior
The study was conducted in healthy adults, with application intended for diabetic foot ulcer patients requiring offloading
What This Means
This research suggests that a new computer algorithm using motion sensors (accelerometers) can accurately track whether patients are wearing their prescribed offloading devices — specifically a removable cast walker on one foot and a special shoe-lift on the other foot — while walking and standing. The study tested 30 healthy adults who wore these devices in a laboratory and then went about their normal lives for 24 hours while keeping a diary of when they wore the devices. The algorithm correctly identified device use more than 99% of the time in the lab, and its results matched participants' own diary records very closely (with less than 3% error) during the real-world monitoring period.
This matters because diabetic foot ulcers are a serious medical complication that require patients to consistently use offloading devices to remove pressure from wounds so they can heal. However, patients often do not wear these devices as prescribed, and until now there has been no reliable, objective way to measure exactly when and how consistently patients use them during daily activities like walking and standing. Previous methods had significant limitations that this new approach was specifically designed to overcome.
This research suggests that this algorithm could be used in future clinical studies to better understand how adherence to offloading devices affects wound healing outcomes. Eventually, it could also be used in clinical settings to help healthcare providers monitor their patients' real-world use of these devices, potentially enabling more personalized interventions for patients who struggle with adherence.
Yalla S, Jhaveri S, Negron-Fernandez N, Crews R, Rosenblatt N. (2026). Development and Validation of an Algorithm to Objectively Quantify Adherence to Offloading Interventions.. Journal of foot and ankle research. https://doi.org/10.1002/jfa2.70170