Validity and Reliability of Two Digital Wound Measurement Tools after Surgery in Patients with Hidradenitis Suppurativa

Hidradenitis suppurativa (HS) is a chronic inflammatory skin disease characterized by recurrent painful inflammatory nodules that can progress to abscesses, tunnels, and scars. These lesions are mainly localized at intertriginous body sites, such as the axillae, groins, anogenital regions, and inframammary folds [1]. The prevalence of HS is estimated to range between 0.4 and 2% in the Western population and has a great negative impact on patient’s health-related quality of life [2, 3]. HS treatment involves multiple pharmacological options, ranging from topical and systemic antibiotics to treatment with biologics. The persistent nature of the disease often requires surgery for local cure of recurrent nodules, abscesses, scarring, and dermal tunnels. Therefore, HS treatment mostly consists of a combination of both pharmacological treatment and surgical interventions [1]. Surgical interventions include incision and drainage, de-roofing, and limited or wide excision [4]. Following surgery, second intention healing is usually preferred to prevent recurrence of the inflammatory lesions [5].

Wound measurement is essential to establish a baseline and can help practitioners monitor the progression of wound healing [6, 7]. Measuring HS wounds is extra challenging because the disease occurs predominantly in the intertriginous body sites, and the surface area of the wounds can be substantial, even up to 450 cm² [7]. Generally, wounds are measured using a ruler, a mathematical method, a tracing method, or digital planimetry [8]. Although the tracing method is still regarded as the most precise option, it is time-consuming and increases the risk of bacterial spreading and contamination after the wound measurement.

Fortunately, digital imaging tools, sometimes using a 3-dimensional (3D) camera, have improved and are increasingly tested for clinical application [9‒11]. Initial studies however measured the wound surface located on flat or mildly convex body surfaces. In the intertriginous body sites, it is difficult to position a wound in a 2-dimensional level. Moreover, due to the concave shape, a difference in depth is present. This difference can be adjusted by using the flexible grid foil but might be overlooked by the 2-dimensional digital camera. A 3D camera can help differentiate the different depths of not only the central wound bed but also the wound borders, which is of importance in wounds localized in the intertriginous body sites [12]. Many previous studies were not sufficiently statistically powered and lacked critical information, such as mentioning of selection criteria, wound size, and wound location, making broad application and acception unlikely [8].

The inSight® 3D device and the ImitoWound app both have specific properties. These properties however have not been studied in concave wounds or intertriginous body sites and are therefore not practicable for measuring wounds on the entire body. For this reason, we aimed to investigate the accuracy, reliability, and feasibility of both the inSight® 3D device and the ImitoWound app for the measurement of the surface area of concave wounds in patients with HS, enabling these tools to be reliably used on all body surfaces.

A prospective observational validation study was conducted between February 2021 and May 2021 in the Department of Dermatology of the Erasmus University Medical Center, Rotterdam. Consecutive patients, aged 18 years or older, with HS, and undergoing limited or wide excision surgery in our department, were screened for inclusion. Patients were excluded if the wound exceeded the 15 × 20 cm surface area of the plastic grid foil. This study was granted exemption from the Dutch Medical Research with Human Subjects Law by the institutional medical ethics review board from the Erasmus MC (MEC-2021-0164).

Three investigators, a physician, a nurse practitioner, and a medical student, each conducted two surface measurements per wound using the inSight® 3D device and the ImitoWound app on two separate moments on the same day according to the users instruction manual. The inSight® 3D device is an application installed on a tablet with a 3D camera and is available for hire through the firm Ekare located in the Netherlands (shown in Fig. 1a). The ImitoWound app is a freely available application trough the Android or Apple store and can be installed on a mobile phone. Calibration markers are needed to perform measurements (shown in Fig. 1b). In this study, both apple and android devices were used.

In order to compare the digital wound measurements with the traditional measurement methods one of the investigators made additional measurements of the wounds. First, the investigator measured the length and width of the wounds with a ruler and additionally calculated the surface area using the following formula: length × width × 0.73 [8, 13]. Second, the tracing method, the gold standard, was applied using Opsite Flexigrid which is a flexible foil containing a metric grid. By outlining and summating the boxes of the grid, the wound surface was calculated.

R statistics (Package “ICC.Sample.Size,” version 1.0, by Rathbone et al. [14], function: calculateIccSampleSize), based on the work of Zou [15], was used to calculate the sample size, including the amount of raters and measurements. For assessment of concurrent validity, we estimated an intraclass correlation coefficient (ICC) of at least 0.95 between the wound measurement tools and the gold standard. Based on literature, an ICC of 0.95 or higher for reliability of both digital measurement tools was assumed [8, 9, 16]. Considering an ICC of 0.9 as null hypothesis [17], a power of 80% and an alpha of 0.05, a sample size of 51 wounds was needed.

Patient characteristics are expressed as the number of patients and percentage (n, %) and mean ± standard deviation or median and interquartile range (IQR) where appropriate. Normality was assessed using the Kolmogorov-Smirnov test. The ICC_2.1 was used for assessment of concurrent validity between the tracing measurement and the digital wound measurement tools [18]. In addition, the ICC_2.1 was used for assessment of intra-rater reliability and inter-rater reliability. The standard error of measurement (SEm) and minimal detectable change (MDC) for intra- and inter-rater reliability of the surface area for the inSight® 3D device and ImitoWound app were calculated [18, 19]. The limits of agreement for the intra- and inter-rater reliability and concurrent validity for the inSight® 3D device ImitoWound app were presented in Bland-Altman plots. The ICC_2.1 between the inSight® 3D device, the ruler method, and the mathematical method and the ICC_2.1 between the ImitoWound app, the ruler method, and mathematical method were calculated [18]. Furthermore, a linear regression analysis was executed to show variability between the tracing method and the two digital wound measurement tools.

A total of 52 wounds were included from 20 patients, who were predominantly female (75%), had a median age of 37.5 (IQR 28.5–50.75), and a median BMI of 31.9 (IQR 27.6–38.3) (see Table 1). Regarding HS severity, 15% of patients were Hurley stage I, 70% Hurley stage II, and 15% Hurley stage III. Wounds were mostly located in the axilla (48.1%) and the inguinal region (25%). The mean surface area of the wounds was 18.7 cm², ranging from 2.50 cm² to 95.75 cm² and a median of 10.1 cm² (IQR 5.8–23.3).

Linear regression analysis showed a positive correlation of 0.980 and 0.974 between the surface assessed using the tracing method and the inSight® 3D device or ImitoWound app, respectively (shown in Fig. 2).

An ICC of 0.987 (95% CI 0.977–0.992) between the inSight® 3D device and the tracing measurement was found. Additionally, an ICC of 0.916 (95% CI 0.857–0.951) and 0.244 (95% CI −0.022 to 0.482) was found between the inSight® 3D device, mathematical method, and length measurement with a ruler, respectively (Table 2).

Using the inSight® 3D device, an ICC of 0.998 (95% CI 0.996–0.999) was found for the intra-rater reliability, resulting in a SEm of 0.95 cm² and a MDC of 2.6 cm². An ICC of 0.997 (95% CI 0.995–0.998) was found for inter-rater reliability, resulting in a SEm of 1.11 cm² and a MDC of 3.1 cm² (Table 2). The results for the intra- and inter-rater differences between raters are shown in online supplementary 1 (for all online suppl. material, see www.karger.com/doi/10.1159/000525844).

After constructing the Bland-Altman plots for the inSight® 3D device, a systematic bias close to zero was found of −0.026 cm² and −0.027 cm², for the intra- and inter-rater reliability, respectively. Furthermore, the systematic bias for the concurrent validity between the tracing measurement and the inSight® 3D device was close to zero with 0.11 cm². The Bland-Altman plots of the inSight® 3D device are shown in Figure 3.

The ICC between the ImitoWound app and the tracing measurement was 0.974 (CI 95% 0.935–0.988). Furthermore, an ICC of 0.901 (CI 95% 0.818–0.945) and an ICC of 0.270 (CI 95% −0.006 to 0.510) were found between the ImitoWound app, the mathematical method, and length measurement with a ruler, respectively (Table 2).

An ICC of 0.978 (95% CI 0.959–0.988) was found using the ImitoWound app for the intra-rater reliability resulting in a SEm of 3.3 cm² and a MDC of 9.9 cm². The inter-rater reliability, in ICC, was 0.964 (95% CI 0.936–0.979) which resulted in a SEm of 3.51 cm² and a MDC of 9.7 cm² (Table 2). The results for the intra- and inter-rater differences between raters are shown in online supplementary 1.

The Bland-Altman plots, as illustrated in Figure 4, showed a systematic bias close to zero for the intra- and inter-rater reliability of the ImitoWound app, with −0.13 cm² and 0.13 cm², respectively. Moreover, the systematic bias for the concurrent validity between the tracing measurement and the ImitoWound app was close to zero with 0.069 cm².

We designed and performed a well-powered validation study with the inSight® 3D device and the ImitoWound app. Both showed excellent results for concurrent validity, intra-rater reliability, and inter-rater reliability compared with the tracing and mathematical method. The excellent correlation with the tracing method is consistent with previous studies validating digital measurement tools [9, 12, 16, 20]. Another validation study used the tablet with 3D camera from Ekare with the inSight® 3D device for wounds located in easily visualized anatomic areas (e.g., not within skin folds). Similar to our study, this study showed excellent results for intra- and inter-rater reliability for the surface measurement of chronic wounds [16]. Additionally, a recent study using the ImitoMeasure app from Imito showed an excellent correlation between the app and the tracing method in pressure ulcers [20]. In contrast to these previous studies, our study was conducted in the intertriginous body sites and still showed an excellent correlation between the digital wound measurements and the gold standard.

An important difference compared with the study using the ImitoMeasure app in pressure ulcers was the size of the wounds. In our study, the wounds were significantly larger with a mean surface area of 18.7 cm² compared with 7.7 cm² in the study using the ImitoMeasure app [20]. Despite this difference in size and in location, similar results were found for the validity and reliability. The ICCs using the ImitoMeasure app were 0.97, 0.99, and 0.98 compared with 0.97, 0.98, and 0.96 using the ImitoWound app, for concurrent validity, intra-rater reliability, and inter-rater reliability, respectively. Our study shows that the inSight® 3D device and the ImitoWound app are also reliable and valid in larger wounds than previously investigated.

Both digital wound measurement tools showed superior results in the intra-rater reliability compared with the inter-rater reliability. This indicates that the wounds should preferably be measured by the same user when conducting several measurements over time. In addition to the quality of the wound measurement, the feasibility for the use of digital wound measurement tools is of significance. For instance, in order to make the most accurate picture of the wound, the inSight® 3D device gives a signal when the optimal distance to the wound is reached, increasing precision and making the device very easy to use. The main disadvantage of the Insight® 3D device is that it can only be acquired by a lease contract with provided tablet, greatly reducing its accessibility for multiple users and for ambulant or home care. This disadvantage is not applicable for the ImitoWound app, which is a freely available app for every Android or Apple device. After applying the calibration marker near the wound edge, a picture is taken with a mobile phone. The calibration marker however could reduce the measurement precision if being applied improperly. Additionally, it slightly reduces the accessibility of the tool as the user needs to have the calibration markers readily available. Overall the ImitoWound app shows superiority to the opinion of the investigators compared with the inSight® 3D device regarding the practical feasibility.

We demonstrated that the inSight® 3D device and the ImitoWound app have distinct properties. The inSight® 3D device is more precise than the ImitoWound app, with a SEm of 0.95 cm² compared with 3.33 cm² in the intra-rater reliability. Furthermore, it shows superiority in the inter-rater reliability with a SEm of 1.11 cm² compared with 3.51 cm². It is noteworthy that a SEm of 3.33 cm² on a mean wound size of 18.7 cm² is not ideal. This means that if the app measures a wound to be 18.7 cm², the true size could range from 15.4 to 22 cm², which is quite a large range and could therefore be an ineffective choice for wounds measured in a clinical research setting. Therefore, we consider the inSight® 3D device more suitable for wound measurement in clinical research or trials, where precise measurement and more accurate reproduction are of key importance. The ImitoWound app however is more suitable for the daily clinical and at home setting for telemedicine purposes, where accessibility and practicality is required.

We believe that the validation of these tools by a physician, a nurse practitioner, and a medical student in parallel is an important strength of this study. Furthermore, this study is well powered and its results generalizable. It enables these digital wound measuring tools to be used in wounds located on all body surfaces, including the difficult, intertriginous sites. Surprisingly, three patients (15%) were Hurley I, meaning that there were no tunnels or scarring in the skin, which is usually treated medically. These patients however presented with persistent, recurrent inflammatory lesions on the same spot. Therefore, these patients were treated surgically. This study was not without limitations. First, we did not validate the 3D option of the inSight® 3D device, which could make a difference by measuring not only the wound surface but also the wound volume. Second, due to the maximum surface area of the used Opsite Flexigrid, not all large wounds could be included leading to a smaller mean surface area.

We anticipate that the ImitoWound app will provide more autonomy to patients in postoperative wound care. Although the app is currently only validated for the use by medical professionals, it has the potential to be used in a home setting where photographs may be taken by home care nurses or relatives. This could facilitate a more accurate follow-up of wounds and wound care at home by means of telemedicine.

This is a well-designed and powered validation study for the measurement of concave wounds. It shows excellent concurrent validity and reliability for wound measurement of both the inSight® 3D device and the ImitoWound app. Both wound measurement tools are proven to be accurate, reliable, and feasible for measuring large postoperative wounds, also on difficult to measure body surfaces. We consider the inSight® 3D device more suitable for clinical research and the ImitoWound app more suitable as a measurement tool in daily clinical practice. Most importantly, the latter app opens the door toward more autonomy for patients receiving wound care at home via telemedicine.

This is the first study to validate digital wound measurement tools to enable use on all body surfaces.

We thank Ekare for providing the inSight® 3D device with account and Imito AG for providing a shared ImitoWound account and calibration stickers.

The Ethical Committee of the Erasmus University Medical Center, Rotterdam, verified that the Medical Research Involving Human Subjects Act (WMO) did not apply to this research (MEC-2021-0164). Written informed consent was obtained from all individual participants included in the study. The authors affirm that human research participants provided informed consent for publication of the images in Figure 1a and b.

Pim Aarts, Johanna C. van Huijstee, Aviël Ragamin, Joanne L. Reeves, Catherine van Montfrans, Hessel H. van der Zee, and Errol P. Prens declare that they have no conflict of interest.

No funding was received from any sponsor and therefore no other party played a role in the preparation of data or the manuscript.

Conceptualization: Errol P. Prens, Hessel H. van der Zee, Catherine van Montfrans, Pim Aarts, Johanna C. van Huijstee, and Aviël Ragamin; data curation: Pim Aarts, Johanna C. van Huijstee, and Joanne L. Reeves; formal analysis: Aviël Ragamin, Pim Aarts, and Johanna C. van Huijstee; investigation: Pim Aarts, Johanna C. van Huijstee, and Errol P. Prens; methodology: Pim Aarts, Johanna C. van Huijstee, and Aviël Ragamin; supervision: Errol P. Prens and Hessel H. van der Zee; writing – original draft: Pim Aarts; writing – review and editing: Errol P. Prens, Hessel H. van der Zee, Catherine van Montfrans, Pim Aarts, Johanna C. van Huijstee, and Aviël Ragamin.

All data generated or analyzed during this study are included in this article and its online supplementary material. Further inquiries can be directed to the corresponding author.