Experiences of using a digital tool, the D-foot, in the screening of risk factors for diabetic foot ulcers

This prospective study was carried out in 2019 in a cohort of patients with diabetes referred to the DPO at Sahlgrenska University Hospital in Gothenburg, Sweden. The purpose of the visit to the DPO was to provide the patient with pressure-relieving footwear, insoles together with shoes, to counteract the occurrence of DFU. The referred patients were contacted by phone and patients interested in participating in the study were informed by the principal investigator and received an invitation letter to the DPO, Fig. 1. The letter included information on how to answer the patient questionnaire in the D-Foot, Additional file 1. The second way for the patients to answer this questionnaire was at the DPO, where the survey was visualised on a tablet (Samsung Galaxy Tab. A 10.1). The patients’ answers were registered before the clinical examination began. If needed, the investigator helped the patients to register their answers on the tablet. The clinical examinations were assessed by one of two CPOs, at the DPO, following the routine in the D-Foot. The daily clinical work process in the prescription is as follows: a) the CPO evaluated the patients’ need for prescribed footwear, b) suggested and discussed a care plan with the patient, c) gave self-care advice and d) provided the patient with footwear as described in clinical guidelines [35]. The information, advice and provision of footwear were based on the patient’s risk of developing DFUs according to the D-Foot examination. Finally, the D-Foot assessment summary was printed out as a PDF report and given to the patient. The footwear was delivered either on the first visit or on a second visit to the DPO, two to 8 weeks later, Additional file 9.

The CPOs documented the assessments, tests, care plan decisions and the information they gave to patients in the local EMR system, Pilot, used at the DPO situated in Region Västra Götaland, Sweden.

The patients’ experiences of the visit to the DPO and the D-Foot examination were evaluated using the Orthotics and Prosthetics Users’ Survey (OPUS) [34]. The survey was filled in after the patients had been provided with their prescribed footwear. The users’ experiences of using the D-Foot were explored by registering the users’ (patients and CPOs) comments as they used the D-Foot at the DPO, Fig. 1. The method used was the think-aloud method, a method frequently used in usability testing where the spontaneous comments from the users are registered, in this case the patients and CPOs [36]. The registration was made by the principal investigator.

The CPOs’ expectations and experiences of using the D-Foot were evaluated using the System Usability Scale (SUS), before and after the study period.

Two experienced CPOs at the DPO at Sahlgrenska University Hospital agreed to participate in the study. They were skilled in prescribing footwear to patients at risk of developing DFUs with assistive devices and had worked as CPOs for nine and 20 years respectively.

Before the start of the study, each CPO completed a D-Foot introductory course which lasted for 2 h on two separate occasions. Moreover, the CPOs participated in a workshop with the purpose of becoming familiar with the D-Foot examination routine.

The D-Foot software has seven functions: 1) a booking system to schedule patients’ appointments, 2) a questionnaire to be answered by patients, 3) examinations to be completed by a healthcare professional, Additional files 2, 4) a system for generating a risk classification for each patient, Additional files 3, 5) a summary as a PDF report, 6) a database containing all registered information and 7) an administration part licensing the CPOs to obtain access to the program [25]. A brief description of some different components included in the D-Foot is also included in Additional file 4 and a summary in Additional file 10. A detailed description has been presented in a previous study by Hellstrand Tang et al. [25]. Based on the results from the previous study, the users (eight CPOs) [25] suggested refinements as compared with the first version. The following improvements were therefore made: 1) adding a test of signs of peripheral angiopathy with the question “Has a healthcare professional confirmed that you have peripheral angiopathy?”, 2) excluding the navicular drop test [37], 3) excluding the test entitled “Can you extend and flex you toes”, 4) excluding the question of whether the foot had areas of excessive pressure with calluses, 5) excluding the assessment “Gait deviation, affected from hip/knee joint” and 6) splitting the assessment of Charcot foot (acute Charcot foot and manifest deformed Charcot foot) [25].

Furthermore, in the current version, the order of the assessment was reorganised to fit more effectively into a more clinical flow with the patient being examined first in a sitting position, followed by standing assessments.

Based on the patients’ answers in the survey and the assessment made by the CPO, a risk classification (1–4) was generated according to the national risk classification system [26, 38], Additional file 3. In this risk classification, the symptoms that are related to each of the risk categories are presented: peripheral neuropathy/angiopathy, foot deformities, skin pathologies, previous DFU/amputation, active DFU/Charcot deformity and/or severe pain syndrome. The risk classification, together with the recommended interventions, such as podiatry, footwear, and regular foot examinations, was automatically displayed on the screen of the tablet and was also included in the PDF report that was given to the patient. Finally, the CPO copied the results from the D-Foot and pasted them into the patient’s EMR and, in addition, all the data were stored in a separate D-Foot database. The reliability and validity of the D-Foot software have previously been presented by Hellstrand et al. [25].

The CPOs recorded their findings on a laptop. Patients answered the patient questionnaire using a smart phone, tablet, or personal computer. A goniometer was used to measure passive maximum dorsal flexion at the metatarsal phalangeal joint and passive dorsiflexion at the ankle joint, as previously described by Hellstrand Tang et al. [25]. A foot calliper was used to measure foot length and foot width, while toe height was measured with a ruler [25], Additional file 4.

The patients answered the Swedish version of the Client Satisfaction with Services module of the OPUS after they had received their footwear. The OPUS module assesses patients’ experiences of the service quality and has been validated with Swedish patients at DPOs [34]. The 10 items are answered on a three-level Likert scale (disagree = 1, agree = 2 and strongly agree = 3) and were analysed by calculating the mean and standard deviation for each question. The OPUS questionnaire is recommended by the Swedish Orthotic and Prosthetic industry advisory council to evaluate patient satisfaction with the services at DPOs located in Sweden [39].

In connection with the start of the study, a Swedish version of the System Usability Scale (SUS) was answered by the CPOs to capture their expectations of using the D-Foot in clinical work, Additional file 5. The SUS consists of 10 questions answered on a five-level Likert scale (from strongly disagree = 1, to agree completely = 5) and is a reliable tool for measuring usability [40]. At the end of the study, the CPOs answered the SUS a second time to assess their experiences of using the D-Foot, Additional file 6. These answers were transformed to a 0–100 scale, where a higher score indicates a higher level of usability.

In addition, at the end of study, the CPOs answered a study-specific questionnaire on how they had experienced the digital foot examination, what information they had given to the patients and how long it took for the CPOs to execute different sequences of the examination, Additional file 7. The questions on time estimations were related to how long it took to: 1) finalise the patient appointment and foot assessment when the D-Foot routine was used (< 30 min., 31–45 min., 46–60 min., > 60 min), 2) make the medical recording based on the D-Foot examination (< 6 min., 6–10 min., 11–15 min., 16–20 min.) and 3) order their prescribed footwear (< 6 min., 6–10 min., 11–15 min., 16–20 min.), Additional file 8.

Descriptive statistics were used to present patient demographics. The mean and standard deviation were calculated for continuous variables. Differences between the included patients and the non-included patients were compared with an independent t-test for age and Pearson’s chi-square test for gender. IBM SPSS Statistics for PC, Version 25, was used for all statistical calculations.

The studied group finally consisted of 90 patients, 53 men and 37 women, with a mean age of 68 ± 13 years. The patients who were not included (n = 17), nine men and eight women, had a mean age of 64 ± 15 years and did not differ significantly from the included participants according to gender (p = 0.65) and age (p = 0.27). Of the entire group, 57 patients had retired and 17 were working, Table 1.

The examination showed that no patient had risk level 1, one (1%) patient had risk level 2, 78 (87%) patients had risk level 3 and 11 (12%) patients had risk level 4. At the examination, the patients were asked about the interventions to prevent DFU they had received. Seventy-eight (87%) patients had had their feet examined by a nurse or doctor during the past year, 59 (66%) answered that they had received podiatry and 59 patients (66%) had been given information about the self-care of their feet.

Eighty-two (91%) of 90 patients answered OPUS. The mean value of the questions ranged from 2.3 to 2.9, where 2 = agree and 3 = strongly agree, Table 2. Satisfaction with services was highest for the questions related to how the CPOs showed courtesy and respect, how the patients were informed about the choice of footwear and the opportunity to express their own concerns. Satisfaction was lowest for the waiting time for the appointment and the co-ordination of services.

One CPO had an SUS score of 77.5 before using the D-Foot and 70.0 after the study was completed. The second CPO had an SUS score of 90.0 before using the D-Foot and 97.5 after the study was completed, Table 3.

In total, 147 comments were registered, (n = 50 from patients and n = 97 from CPOs), Table 4. Overall, 55 comments were related to improvements (n = 26 from patients and n = 29 from CPOs). Patients suggested refinements to the patient questionnaire regarding the perceptions of neuropathy; pain, previous ulcers, sweat and HbA1c. Patients found questions such as “Have you had a foot ulcer? and “Are your feet less sweaty now compared with recent years?” difficult to answer. Patients suggested specifying the question of “Have you previously had hard-to-heal ulcers?” and removing the question about sweaty feet, because some patients said that they had never had sweaty feet. Further, they suggested expanding the question on the sensation of tingling or numbness to include sensations of cramp, walking on pillows and feelings such as “it feels like a band around my foot”. The question about the presence of pain should be more specific, such as “How much pain have you experienced in your feet during the past week”.

Other suggested improvements, coming from both patients and CPOs, were to design the PDF report in an easy-to-understand format with larger text, linking the risk grade to each of the specific risk factors and, in addition, creating customised information and advice about self-care of the feet for patients. Requests were also made to be able to post the report in digital format in addition to the printable version. The individual risk factors should be linked to the recommended treatment (e.g. recommendation to use a high toe box in the presence of hammer toes). Moreover, digital referrals to other healthcare professionals, based on the results, should be included in upcoming versions.

The CPOs requested integration with co-working systems, such as the population register, the EMR used by other healthcare professionals and the EMR used at the DPO. Loss of wi-fi, breakdown of the EMR at the DPO and problems registering findings in the D-Foot system were registered. Eleven patients were not booked in the D-Foot system and these patients were therefore unable to answer the patients’ questions at home, prior to the visit.

The CPOs felt that they had provided “complete” information regarding a) how patients should perform self-care for their feet, b) where patients could obtain additional information, c) potential risks when using the assistive device and d) warning signs relating to health and the assistive device. Regarding the questions: a) whether the examination took place in privacy, b) information on the D-Foot results and c) whether the patient had received oral and written information about the assistive device, they were judged to have been performed “completely”. Two questions were judged to be less well fulfilled by one of the CPOs: a) the discussion of the health condition and b) whether the results of the D-Foot survey were explained in a way the patient understood, Table 5.

The estimated time required for different sequences when meeting the patients was assessed equally by the two CPOs. The time required for the patient appointment was 40–60 minutes. Writing the medical record took six to 10 minutes and ordering shoes and materials took six to 10 minutes. The patient’s visit was finalised within a total of 52–80 minutes.