Beruflich Dokumente
Kultur Dokumente
Reprint requests:
ABSTRACT
Chan Yeong Heo, Department of Plastic
and Reconstructive Surgery, Seoul National We investigated the accuracy of pressure injury evaluation using tele-devices and
University College of Medicine, examined the concordance between automatically generated recommendations and
103 Daehak-ro (Yeongeon-dong), Jongno- primary manual recommendations. Caregivers took photos and videos of pressure
gu, Seoul 03080, Republic of Korea.
injuries using smartphones with built-in cameras and uploaded the media to the
Tel: +(82) 31-787-7222;
Fax: +(82) 31-787-4055;
application. The wound team evaluated the wound using a specially modified ver-
Email: lionheo@snu.ac.kr sion of the Pressure Sore Status Tool. This was compared with the Pressure Sore
Status Tool score assessed during the actual examination of the patient. We devel-
Manuscript received: April 3, 2017
oped an automatic algorithm for dressing based on the Pressure Sore Status Tool
Accepted in final form: August 27, 2018
score, checking for consistency between this and the primary manual recommenda-
tion. A total of 60 patients diagnosed with pressure injuries were included. The κ
coefficients indicated substantial agreement for wound size and total score, and
DOI:10.1111/wrr.2
excellent for all other items. We found that the overall concordance rates were sta-
Jae Hoon Jeong and Chan Yeong Heo tistically significant for all items (p < 0.001). For the primary dressing, the κ coeffi-
contributed equally to this article as cient for the concordance rate of automatic algorithm and manual recommendation
corresponding authors. was 0.771, while that of teleconsultation system and manual recommendation was
0.971. For the secondary dressing, the figures were 0.798 and 0.989, respectively.
All values were statistically significant (p < 0.001). We presented strong evidence
documenting the utilization of a smartphone, patient-driven system, and demon-
strated that the measurements obtained were comparable to the ones obtained by a
trained, on-site, wound team. Furthermore, we confirmed agreement between auto-
matically generated recommendations and primary manual recommendations.
Wound Rep Reg (2018) 26 S19–S26 © 2018 by the Wound Healing Society S19
Teleconsultation of chronic wound management Pak et al.
result in poor quality of life for both patients and their care- addition, we chose the iPhone, which offered the best secu-
givers. It is therefore crucial to treat patients with pressure rity available at the time, and designed the system to delete
injuries appropriately.15 When monitoring the clinical course the data from the mobile device following transmission.
of patients with chronic pressure injuries, it is important to Data were collected, monitored, and analyzed, and feed-
achieve prompt recovery from symptoms, thereby minimiz- back was obtained in accordance with the standards of the
ing the need for in-hospital intervention. National Pressure Ulcer Advisory Panel and its Korean
Consequently, we developed chronic pressure injury tele- equivalent, the Bundang Pressure Ulcer Advisory Panel. The
consultation methods based on a smartphone store-and- feasibility of the system was assessed through the collabora-
forward system. Using information that patients and care- tion of a Wound, Ostomy, and Continence (WOC) nurse, a
givers sent, a doctor would make a diagnosis remotely. In plastic surgeon, and a researcher.
addition, the system was capable of automatically recom-
mending a diagnosis-based treatment. We investigated the Data collection (mobile device)
accuracy of the pressure injury evaluation via tele-device, and
examined the agreement between the automatically generated Pressure injuries are commonly diagnosed through a gross
recommendation and the primary manual recommendation. examination, for which the data can be collected simply by
using the built-in camera of a smartphone. For the current
study, we used the standard touch screen user interface and
METHODS camera installed in the iPhone 3GS and iPhone 4, with iOS
3.0 and iOS 5.0 operating systems, respectively.
Study population (1) Measurement of pressure injury size
Patients undergoing outpatient treatment were selected from Caregivers of patients were instructed to remove the
a group of patients diagnosed with pressure injuries in the wound dressings and expose the wound sites for photo-
plastic surgery clinic. Only patients who were 21 years of graphs. At this time, attempts were made to examine the
age or older with a wound less than 100 cm2 were included. amount of exudate at the wound sites. Caregivers were then
The following patients were excluded: patients diagnosed
with osteomyelitis who had not yet been treated, patients
with fistulae, patients who had participated in other drug
studies within a month of this study, and patients considered
by the researchers to present difficulties for participation in a
clinical trial.
Study oversight
This study was a single-institution trial to test the accuracy
of pressure injury evaluation via tele-device, and to examine
the agreement between the automatically generated recom-
mendation and the primary manual recommendation. Each
patient visited a hospital once a week for 4 weeks; a total of
four evaluations were made per patient. The current study
was approved by the Institutional Review Board (IRB) of
Seoul National University Bundang Hospital (IRB # E-
1006-050-001, E-1106/063–005, and E-1206/158–001), and
all patients and legal guardians provided written informed
consent to participate.
Study design
System protocol
Our system followed a design-implementation-evaluation
developmental cycle. Design and implementation were
based on data collection, monitoring, analysis, and feed-
back.16 The system was designed to measure and process
biometric data, and to detect changes in biological markers Figure 1. System diagram. The staff portal/ OS terminal on
associated with the health status of patients, using a smart- the first line is the cyber space where the users upload infor-
phone interface. In addition, it alerted patients to changes in mation. In the second telemedicine server, the API layer
their health status and transmitted relevant information to serves to receive user-uploaded information. The control layer
physicians (Figure 1). This study linked the electronic medi- analyzes and processes user-uploaded information. The data
cal records of Seoul National University Bundang Hospital access layer allows the processed information to be stored in
network, and the information was stored on the hospital’s the data center. The third database is a space where various
servers. As a result, the data received the same level of pro- data are stored. OS, operating system; HTTPS, hypertext
tection as the other medical records. Not only did we transfer protocol over secure socket layer; XML, extensible
enhance the security of the application itself, but also we markup language; API, application programming interface;
prevented developers from accessing patient information. In WAS, web application server; DBC, database connectivity.
S20 Wound Rep Reg (2018) 26 S19–S26 © 2018 by the Wound Healing Society
Pak et al. Teleconsultation of chronic wound management
instructed to transmit the photographs. The size of the pres- via an iPhone touch screen interface. Additionally, patients
sure injuries was automatically measured using a 1 cm × 1 cm could obtain further information from the researcher using
indicator on the screen. Automatic size measurement was an open question format.
performed by touching the screen and dragging or dropping
the indicator, thereby modifying its movement and eventu-
Remote physician evaluation and recommendation
ally adjusting the recorded size of the pressure injury
(Figure 2a). The resolution of the iPhone 3GS camera sen- Patient data were transmitted from the smartphone to a web-
sors was set to 3,000,000, while the resolution of the iPhone site via an open application programming interface (API).
4 camera sensors was set to 5,000,000 pixels. From the website, physicians monitored baseline characteris-
(2) Videography tics, prescription and treatment data, and types of dressing.
It is impossible to grossly examine the depth and direction (1) Wound evaluation
of wounds using two-dimensional images, and difficult to suf- One plastic surgeon, one WOC nurse, and one researcher
ficiently examine a pressure injury using a single still image. grossly examined the wound sites of each patient and
Clean swabs were therefore used to sweep around the edges assessed the pressure injuries. We evaluated the wounds
of the deepest point of the pressure injury in each direction using a specially modified version of the Pressure Sore Sta-
for the purposes of presenting the depth of the wound on the tus Tool (PSST), which comprised 13 questions; in an
video. The function of the video or the system was designed attempt to obtain the wound scores (Table 1). Subsequently,
to run and exit the program using the touch screen user inter- caregivers transmitted images, videos, and other relevant
face, for which the video settings of the iPhone 3GS and information concerning the pressure injury to the website
iPhone 4 were optimized at 30 frames per second at a resolu- via a smartphone (Figure 2b). The same information was
tion of 640 × 480 and 640 × 940 pixels, respectively. also collected through teleconsultation, which was used for
(3) Medical examination comparison. The evaluators were blinded to all information
The degree of recovery from a pressure injury is depen- other than wound-related data in order to minimize bias in
dent on baseline characteristics, and different types of dress- determining the concordance rate.
ing are recommended based on infectious symptoms.17 We (2) Automatically generated recommendation
used a patient-reported questionnaire and caregiver informa- Based on PSST scores and algorithm design, specific
tion to collect patient data regarding the presence of infec- dressing types (primary dressing for packing dead space,
tious symptoms. These baseline characteristics were secondary dressing for covering) were recommended and
collected during the initial registration and then saved as part transmitted to patients. To ensure patient safety, the recom-
of each user’s profile. Data were entered via an onscreen mendations were reviewed and approved by physicians. If
keyboard or voice recording, the latter being an option for applicable, physicians forwarded the recommendations to
elderly patients who are often inexperienced in entering text patients after making modifications through a click reaction.
Figure 2. Data collection and physician evaluation. (A) Caregivers of patients were instructed to transmit the photographs. The
size of the pressure injuries was automatically measured by a 1 cm × 1 cm indicator on the screen. Automatic size measure-
ment was performed by touching the screen and dragging or dropping the indicator, thereby modifying its movement and even-
tually adjusting the recorded size of the pressure injury. (B) The wound team checked all PSST items based on the information
the caregiver sent. Medical staff who wished to confirm the explanation could obtain information by clicking on the separate link
window labeled “help.” PSST, Pressure Sore Status Tool.
Wound Rep Reg (2018) 26 S19–S26 © 2018 by the Wound Healing Society S21
Teleconsultation of chronic wound management Pak et al.
Items Criteria
(Continues)
S22 Wound Rep Reg (2018) 26 S19–S26 © 2018 by the Wound Healing Society
Pak et al. Teleconsultation of chronic wound management
Table 1. (Continued)
Items Criteria
Wound Rep Reg (2018) 26 S19–S26 © 2018 by the Wound Healing Society S23
Teleconsultation of chronic wound management Pak et al.
Primary dressing Automatic algorithm vs. Manual 0.771 (p < 0.001) 84.0
recommendation
Teleconsultation system vs. Manual 0.971 (p < 0.001) 98.1
recommendation
Secondary dressing Automatic algorithm vs. Manual 0.798 (p < 0.001) 92.6
recommendation
Teleconsultation system vs. Manual 0.989 (p < 0.001) 99.6
recommendation
S24 Wound Rep Reg (2018) 26 S19–S26 © 2018 by the Wound Healing Society
Pak et al. Teleconsultation of chronic wound management
designed to remind patients to adjust their position at 2 hour matching rate of virtual and on-site PSST. Further studies
intervals to prevent the aggravation of pressure injuries. It are required to design a system that can make decisions
also provides information on prescription data, wound dress- based on a combination of various data, rather than virtual
ing methods, and medical examination results in text form. images alone. Second, we unfortunately could not measure
In addition, it allows patients to access information in a the exact cost of the system in this study, because the medi-
diary format by date, in which the latest recommendation cal fee for telemedicine had not been established in our
for dressing materials appears at the beginning. Our system country to provide evidence of cost-effectiveness. In Korea,
also incorporates nutritional data to help develop patient- it usually costs about $200 to travel to the hospital four
specific knowledge based on the Clinical Decision Support times, so there is likely to be a certain degree of cost sav-
System. Patients feel comfortable when they are aware that ings. Third, we could not apply this system to all chronic
their clinical course is being monitored meticulously and wounds. It works well when applied to stable chronic
continuously in a teleconsultation setting, which in turn may wounds, but it is difficult to use when the wound necessi-
increase their adoption of the system.15 tates urgent care, or emergent interventions are required due
This system has several strengths. First, mobile devices to aggravation or sudden worsening. Fourth, there may be a
are widely used, so patients and caregivers do not have to bias depending on education level, familiarity with smart-
buy extra equipment. Patients can save a great deal of phones, and the clinical conditions of caregivers. There is a
money and time. It is possible to eliminate problems that specificity of telemedicine that involves the interaction
may occur when moving. Most importantly, the degree of between the patient and the physician using the devices. We
user convenience is relatively high among caregivers who attempted to simplify the procedures that caregivers must
are familiar with currently available smartphones. Second, in follow for the transmission of images, video, and other data
this study, the virtual examination was designed to be as to physicians, and made size measurement automatic. We
accurate as possible through the use of both images and think this bias can be reduced through caregiver training and
video of the wounds. It is difficult to sufficiently examine increased efforts to make our system more user-friendly.
pressure injuries using still images alone. Third, the differ- Further large-scale prospective studies are warranted to
ence between actual size and measurement was minimized support our results, in conjunction with greater efforts to
according to users, by a tool that could objectively measure improve the convenience of our system and to provide an
wound size. evidence-based rationale. In addition, more efforts are
The authors recommend that telemedicine researchers needed to confirm the scope of our system beyond its rate of
address security concerns to increase the availability of tele- concordance. Our innovative system can contribute to mini-
medicine services and improve the quality of care pro- mizing the likelihood of inefficient, unprofessional wound
vided.19,20 Security issues should be eliminated at every management in nonhospital settings, achieving prompt
stage: database server, application, and mobile device. It is recovery from pressure injuries, and reducing medical costs
important to store data on a secure hospital server and the for patients. As such, it may provide a foundation for the
data should receive protection that is at least equal to or future development of teleconsultation in clinical practice.
higher than that of the other medical records. From the
beginning of the application’s development, researchers
must collaborate with researchers in security and associated
ACKNOWLEDGMENTS
fields. In this process, we prevented even developers from This work was supported by KT and the Seoul National
accessing medical information. Nevertheless, there are limits University Bundang Hospital u-Health cooperative research
to the security of mobile devices, because they use machine- project.
based security systems; if the mobile device is hacked, pro-
tection may not be possible. For this reason, researchers
should choose the device that offers the best security avail-
REFERENCES
able. In addition, we recommended designing the system to 1. Wald HS, Dube CE, Anthony DC. Untangling the web--the
delete the data from the mobile device following transmis- impact of internet use on health care and the physician-patient
sion (store-and-forward system). We are in full agreement relationship. Patient Educ Couns 2007; 68: 218–24.
with concerns about security, the most important obstacle to 2. Michael BD, Geleta D. Development of ClickClinica: a novel
the development of telemedicine, and must strive for ade- smartphone application to generate real-time global disease sur-
quate protection in the future. veillance and clinical practice data. BMC Med Inform Decis
There is an argument related to the cost of telemedicine. Mak 2013; 13: 70.
Costs vary greatly depending on the medical environment of 3. Acampora G, Cook DJ, Rashidi P, Vasilakos AV. A survey on
a given country. Cost may have an impact on the effective- ambient intelligence in health care. Proc IEEE Inst Electr Elec-
ness of teleconsultation. For telemedicine to be widely used, tron Eng 2013; 101: 2470–94.
appropriate medical fees must be set up through social 4. O’Connor PJ, Desai JR, Butler JC, Kharbanda EO,
agreement. Sperl-Hillen JM. Current status and future prospects for elec-
There are several limitations to this study. First, diagnos- tronic point-of-care clinical decision support in diabetes care.
ing and treating wounds solely on the basis of a virtual Curr Diab Rep 2013; 13: 172–6.
examination is not easy, even though we did the best we 5. Trovato M, Scholer A, Vallejo E, Buncke G, Granick M. eConsul-
could with 2D images. If the camera module is developed, it tation in plastic and reconstructive surgery. Eplasty 2011; 11: e48.
will serve to complement this part. The 3D image analysis 6. Jones SM, Banwell PE, Shakespeare PG. Telemedicine in
technology with scenes, for example, in which the therapist wound healing. Int Wound J 2004; 1: 225–30.
presses the surrounding tissues when shooting the video, 7. Hazenberg C, Van Baal J, Manning E, Bril A, Bus SA. The
could be used to reduce the score difference and increase the validity and reliability of diagnosing foot ulcers and
Wound Rep Reg (2018) 26 S19–S26 © 2018 by the Wound Healing Society S25
Teleconsultation of chronic wound management Pak et al.
pre-ulcerative lesions in diabetes using advanced digital pho- 14. Tachakra S. Level of diagnostic confidence, accuracy, and rea-
tography. Diabetes Technol Ther 2010; 12: 1011–7. sons for mistakes in teleradiology for minor injuries. Telemed J
8. Coiera E. Communication systems in healthcare. Clin Biochem E Health 2002; 8: 111–21.
Rev 2006; 27: 89–98. 15. Sukumar S, Roghmann F, Trinh VQ, Sammon JD, Gervais M-K,
9. Barnard CM, Goldyne ME. Evaluation of an asynchronous Tan H-J, et al. National trends in hospital-acquired preventable
tele-consultation system for diagnosis of skin cancer and other adverse events after major cancer surgery in the USA. BMJ Open
skin diseases. Telemed J E Health 2000; 6: 379–84. 2013; 3: e002843.
10. Eminovic N, De Keizer N, Bindels P, Hasman A. Maturity of 16. Helal A, Cook DJ, Schmalz M. Smart home-based health plat-
teledermatology evaluation research: a systematic literature form for behavioral monitoring and alteration of diabetes
review. Br J Dermatol 2007; 156: 412–9. patients. J Diabetes Sci Technol 2009; 3: 141–8.
11. Du Moulin M, Bullens-Goessens Y, Henquet C, Brunenberg 17. Bowler PG, Duerden BI, Armstrong DG. Wound microbiology
D, de Bruyn-Geraerds D, Winkens R, et al. The reliability of and associated approaches to wound management. Clin Micro-
diagnosis using store-and-forward teledermatology. J Telemed biol Rev 2001; 14: 244–69.
Telecare 2003; 9: 249–52. 18. Landis JR, Koch GG. The measurement of observer agreement
12. Salmhofer W, Hofmann-Wellenhof R, Gabler G, Rieger- for categorical data. Biometrics 1977; 33: 159–74.
Engelbogen K, Gunegger D, Binder B, et al. Wound teleconsulta- 19. Garg V, Brewer J. Telemedicine security: a systematic review.
tion in patients with chronic leg ulcers. Dermatology 2005; 210: J Diabetes Sci Technol 2011; 5: 768–77.
211–7. 20. Barlow J, Singh D, Bayer S, Curry R. A systematic review of
13. Mair F, Whitten P. Systematic review of studies of the benefits of home telecare for frail elderly people and those
patient satisfaction with telemedicine. BMJ 2000; 320: with long-term conditions. J Telemed Telecare 2007; 13:
1517–20. 172–9.
S26 Wound Rep Reg (2018) 26 S19–S26 © 2018 by the Wound Healing Society