PII: S0741-5214(98)70011-4 Telemedicine in vascular surgery: Feasibility o f digital imaging for remote management of wounds D o u g l a s J. W i r t h l i n , M D , S y a m B u r a d a g u n t a , B A , R o g e r A. E d w a r d s , S c D , D a v i d C . B r e w s t e r , M D , R i c h a r d P. C a m b r i a , M D , J o n a t h a n P. G e r t l e r , M D , G l e n n M . L a M u r a g l i a , M D , D i a n e E . J o r d a n , B S , J o s e p h C. K v e d a r , M D , a n d W i l l i a m M . A b b o t t , M D , Boston, Mass. Purpose: Telemedicine coupled w i t h digital p h o t o g r a p h y could potentially improve the quality o f o u t p a t i e n t w o u n d care and decrease medical cost b y allowing h o m e care nurs- es t o electronically t r a n s m i t images o f patients' wounds t o treating surgeons. T o deter- mine the feasibility o f this technology, we c o m p a r e d bedside w o u n d examination b y onsite surgeons w i t h viewing digital images o f wounds b y r e m o t e surgeons. Methods: Over 6 weeks, 38 wounds in 24 inpatients were p h o t o g r a p h e d with a K o d a k D C 5 0 digital camera (resolution 756 × 504 pixels/in2). Agreements regarding w o u n d description (edema, erythema, cellulitis, necrosis, gangrene, ischemia, and granulation) and w o u n d m a n a g e m e n t (presence o f healing problems, need for emergent evaluation, need f o r antibiotics, and need for hospitalization) were calculated a m o n g onsite sur- geons a n d between onsite and r e m o t e surgeons. Sensitivity and specificity o f r e m o t e w o u n d diagnosis c o m p a r e d w i t h bedside examination were calculated. Potential corre- lates o f agreement, level o f surgical training, certainty o f diagnosis, and w o u n d type were evaluated b y multivariate analysis. Results:Agreement between onsite and r e m o t e surgeons (66% t o 95% for w o u n d descrip- tion and 64% to 95% for w o u n d management) matched agreement a m o n g onsite sur- geons (64% t o 85% f o r w o u n d description and 63% to 91% f o r w o u n d management). Moreover, w h e n onsite agreement was low (i.e., 64% f o r erythema) agreement between onsite and r e m o t e surgeons was similarly low (i.e., 66% f o r erythema). Sensitivity o f r e m o t e diagnosis ranged f r o m 78% (gangrene) to 98% (presence o f w o u n d healing p r o b - lem), whereas specificity ranged f r o m 27% (erythema) to 100% (ischemia). Agreement was influenced b y w o u n d type (p < 0.01) b u t n o t b y certainty o f diagnosis (p > 0.01) o r level o f surgical training (p > 0.01). Conclusions: Wound evaluation on the basis o f viewing digital images is comparable with standard w o u n d examination and renders similar diagnoses and treatment in the majority o f cases. Digital imaging for remote w o u n d management is feasible and holds significant promise for improving outpatient vascular w o u n d care. (J Vasc Surg 1998;27:1089 - 1100.) From the Department of Surgery (Drs. Wirthlin, Brewster, Cambria, Gertler, LaMuraglia, and Abbott), and the Partners Telemedicine Center (S. Buradagunta, D. E. Jordan, and Drs. Edwards and Kvedar), Massachusetts General Hospital; and Harvard Medical School. Presented at the Twenty-fourth Annual Meeting of the New England Society for Vascular Surgery, Bolton Landing, N.Y., Sep. 18-19, 1997. Reprint requests: Douglas J. Wirthlin, MD, Department of Surgery/Division of Vascular Surgery, Massachusetts General Hospital, WAC 458, Boston, MA 02114. Copyright © 1998 by The Society for Vascular Surgery and International Society for Cardiovascular Surgery, North American Chapter. 0741-5214/98/$5.00 + 0 2 4 / 6 / 8 9 6 1 6 Telemedicine is an e v o M n g field t h a t c o m b i n e s t e l e c o m m u n i c a t i o n s a n d i n f o r m a t i o n technologies to p r o v i d e r e m o t e medical care and ranges f r o m activi- ties as simple as t e l e p h o n e consultation t o t e c h n o l o - gy as c o m p l e x as telesurgery.1 Electronic transmission o f clinical images for r e m o t e consultation is o n e c o m - p o n e n t o f t e l e m e d i c i n e t h a t has b e e n successfully i m p l e m e n t e d a n d tested in the fields o f d e r m a t o l o g y , pathology, and radiology. 2-8 M o r e o v e r , Kvedar et al.9 o b s e r v e d t h a t as m a n y as 83% o f d e r m a t o l o g i c diag- noses c o u l d accurately be m a d e by viewing still digi- tal images. Accordingly, we h y p o t h e s i z e d t h a t still digital i m a g e s c o u l d precisely r e p r e s e n t vascular 1089 JOURNAL OF VASCULAR SURGERY 1090 Wirthlin et al. June 1998 surgery wounds and that w o u n d evaluation and man- agement based on viewing digital images w o u l d closely match that o f evaluation and management based on bedside examination. Validation o f this hypothesis would suggest that digital imaging cou- pled with telemedicine could be implemented to enhance outpatient w o u n d care. For example, using current technology, visiting nurses could photograph patients' wounds and transmit the digital images over a short period o f time to the treating surgeon, thus allowing surgeons to manage w o u n d s remotely. Remote w o u n d management, in turn, has the poten- tial to decrease the frequency o f office visits, prevent unnecessary "urgent" w o u n d evaluations, and short- en hospital stay for patients with w o u n d complica- tions. This potential application o f telemedicine is par- ticularly appealing because o f the frequency and chronic nature o f w o u n d complications in vascular surgery, the large consumption o f resources to trcat these wounds, and the shift o f health care from the hospital to the outpatient setting. Johnson et al.10 observed that 40% to 50% o f patients who undergo lower extremity bypass procedures have a nonheal- ing ulcer that may require several months to heal despite a functioning bypass graft. Also, o f those patients who undergo lower extremity bypass proce- dures, 20% to 30% will have a w o u n d complication that will lengthen hospitalization and necessitate prolonged w o u n d care. l i d 3 Moreover, as suggested by Calligaro and others w h o have evaluated the impact o f care pathways in vascular surgery, there is increasing pressure to shorten hospitalization and decrease the cost o f care.14,15 Thus the purpose o f this study is to determine the accuracy o f w o u n d evaluation and management based on viewing digi- tal images as an initial step in testing the feasibility o f implementing digital imaging for r e m o t e w o u n d management. P A T I E N T S A N D M E T H O D S The feasibility o f digital imaging for remote w o u n d diagnosis was evaluated by comparing w o u n d evaluation and management based on viewing digital images o f wounds with evaluation and management based on examining wounds at the bedside. Surgeons who examined wounds directly were labeled onsite surgeons, and surgeons w h o viewed digital images were labeled remote surgeons. The onsite surgeons' w o u n d evaluation and management constituted the " g o l d standard" in this study. Fig. 1 shows an overview o f the study design. The experimental pro- tocol was approved by the Institutional Review Board o f Massachusctts General Hospital (Accession #9607618), and all subjects gave informed consent. P a t i e n t s . Vascular surgery inpatients at the Massachusetts General Hospital from March 18, 1996, to May 5, 1996, were invited to participate. Eligible patients included those inpatients who had recently undergone a lower extremity bypass proce- dure or amputation or were admitted for a w o u n d healing problem (nonhealing ulcer, necrotic/gan- grenous toes, cellulitis). Approximately 36% o f total eligible patients were included in this study, and eli- gible patients who did not participate were excluded as a result o f factors beyond the control o f the study: busy clinical activity, inability to coordinate digital imaging with onsite w o u n d evaluation, and refusal to consent. Twenty-four inpatients (six female and 18 male) with 38 separate wounds were pho- tographed. The wounds were categorized as follows: postoperative incision (n = 16), amputation site (n = 3), n e c r o t i c / g a n g r e n o u s toes (n = 11), and non- healing ulcer (n = 8). The number o f wounds per extremity and w o u n d category were assigned on the basis o f the onsite surgeons' examination and chart review. For postoperative incisions, any area o f the incision with a w o u n d complication that might inde- pendently influence w o u n d evaluation was consid- ered a separate wound. A brief history (age, sex, reason and date o f admission, medical and surgical history, vascular physical examination, and treatments rendered from admission to the time o f imaging) on each patient was obtained by chart review. D i g i t a l p h o t o g r a p h y a n d i m a g e display. During morning rounds, a nonphysician w i t h o u t any formal p h o t o g r a p h y training (S. B.) pho- tographed all wounds using a digital camera (Kodak D C 50, Eastman Kodak Co., Rochester, N.Y.; reso- lution 736 x 504 pixels/in 2, single C C D chip, 24- bit color). Table I describes our imaging protocol. Seventeen o f the 38 wounds were photographed on more than one day, rendering 45 image sets for comparison and 183 total images. Images were stored as highest-quality J P E G 16 files (Joint Photographic Experts Group, a compression algo- rithm for digital images) and were converted t o a Microsoft P o w e r p o i n t slide presentation to be viewed on a computer monitor (Mitsubishi 91TXM at the maximum attainable resolution). Fig. 2 shows examples o f w o u n d images. All images were graded by a nonsurgeon, non- medical photographer (E. 1~ M.) using a rating sys- tem developed to grade the image quality o f stan- dard 35 mm photographs. 9 Images were assigned a JOURNAL OF VASCULAR SURGERY Volume 27, Number 6 W i r t h l i n et al. 1091 I II 2 to 4 surgeons examine wounds at the bedside at the time of digital imaging, I I II II II I I 2 to 4 surgeons evaluate wounds by viewing digital images of wounds at a later time. I Wound questionnaire I I I Fig. 1. Study design. Using a standard questionnaire and bedside wound examination as the "gold standard," the feasibility of digital imaging for remote wound management was mea- sured by concordance between bedside and remote surgeons. p h o t o g r a p h i c quality r a t i n g (0 [ u n r e a d a b l e ] t o 5 [perfect quality]). Using this grading system, a score o f 4 is highest quality rating attainable with o u r dig- ital c a m e r a b e c a u s e t h e r e s o l u t i o n , c o l o r r a n g e , image m e m o r y , and lighting are inferior to standard 35 m m p h o t o g r a p h s . H o w e v e r , there are a n u m b e r o f " h i g h e r - e n d " digital cameras with p h o t o g r a p h i c features and characteristics m o r e similar to standard 35 m m cameras that w o u l d have scored consistently higher t h a n the camera used in this study. Wound e v a l u a t i o n . Wounds were evaluated by two to f o u r onsite surgeons and two t o four r e m o t e surgeons. Table II shows the n u m b e r o f evaluations c o m p l e t e d and the level o f surgical training o f t h e evaluators (five vascular attendings, five vascular fel- lows, a n d six surgical residents). O n s i t e s u r g e o n s examined each w o u n d at the bedside near the time o f digital imaging and had n o restrictions in m e t h o d o f e x a m i n a t i o n , fo r example, p al p at i o n , olfaction, viewing f r o m multiple angles, and time o f evalua- t i o n . T h e s e t t i n g ( l i g h t i n g , p a t i e n t p o s i t i o n , a n d b a c k g r o u n d ) d u r i n g o n si t e e x a m i n a t i o n m a t c h e d t h e s e t t i n g d u r i n g digital p h o t o g r a p h y . I n m o s t cases, onsite surgeons were involved in the care o f t h e patient, an d occasionally w h e n an onsite s u r g e o n was n o t t h e t r e a t i n g s u r g e o n a b r i e f h i s t o r y was given before examining t h e w o u n d . R e m o t e surgeons evaluated w o u n d s by viewing images o n a c o m p u t e r m o n i t o r some time after dig- ital imaging. Before viewing w o u n d images, r e m o t e surgeons were given a b ri ef medical h i st o ry o f each patient. T h e r e was n o time limit for viewing images and r e m o t e surgeons were allowed t o scroll back and f o r t h a m o n g d i f f e r e n t views o f t h e w o u n d s . N o image m a n i p u l a t i o n s such as m ag n i fi cat i o n , c o l o r e n h a n c e m e n t , o r contrast e n h a n c e m e n t were used. JOURNAL OF VASCULAR SURGERY I 0 9 2 Wirthlin ctal. June 1998 T a b l e I . I m a g i n g p r o t o c o l Camera (Kodak DC50) settings Auto focus, auto flash, auto exposure Lighting Window shade closed, overhead examination light on Position of patient Supine Presentation of wound Blue pad under extremity Transparent ruler and patient identifier number placed adja- cent to wound Photographs taken (image set) View of entire extremity (camera 40 inches from wound) View of thigh (camera 20 inches from wound) View of leg and foot (camera 20 inches from wound) Close-up views* (camera 18 inches from wound with lens set on telephoto) *Close-up views were taken of specific wounds (gan- grenous/necrotic toes or nonhealing ulcers, and areas along a postoperative incision with a wound complication). The compo- sition and number of close-up views taken per extremity were determined by an onsite surgeon. T a b l e I I . W o u n d evaluations Onsite Remote Attending 7 55 Fellow 76 65 Resident* 16 0 Total 99 i20 *Only two of six residents had completed less than 4 years of training. After evaluating w o u n d s , b o t h onsite a n d r e m o t e s u r g e o n s c o m p l e t e d a s t a n d a r d w o u n d q u e s t i o n - naire, w h i c h addressed questions r e g a r d i n g w o u n d d e s c r i p t i o n a n d w o u n d m a n a g e m e n t . S u r g e o n s answered either " y e s , " " n o , " " p r e s e n t , " o r " n o t p r e - s e n t " t o each q u e s t i o n a n d r e c o r d e d a level o f cer- t a i n t y f o r e a c h r e s p o n s e (1 [ n o t c e r t a i n ] t o 10 [absolutely certain]; T a b l e I I I ) . S t a t i s t i c a l m e t h o d s . All d a t a w e r e c o l l e c t e d , s t o r e d in M i c r o s o f t Excel files, a n d i m p o r t e d i n t o SAS (SAS I n s t i t u t e I n c . , Cary, N . C . ) f o r statistical analysis. Descriptive analysis was p e r f o r m e d for the f o l l o w i n g e n d p o i n t s : i m a g e q u a l i t y ( m e a n a n d range), certainty o f r e s p o n s e ( m e a n a n d range), a n d a g r e e m e n t a m o n g a n d b e t w e e n s u r g e o n g r o u p s ( m e a n p e r c e n t a g r e e m e n t a n d k a p p a values for each w o u n d d e s c r i p t o r a n d m a n a g e m e n t d e c i s i o n ) . I n a d d i t i o n , p r e v a l e n c e o f a g r e e m e n t s a n d d i s a g r e e - m e n t s a l o n g w i t h t w o v a r i a t i o n s o f k a p p a values, k a p p a ( n o r ) a n d k a p p a ( m a x ) 17 w e r e calculated (scc Appendix). I n an a t t e m p t t o c o n t r o l for s u r g e o n vari- a t i o n , a g r e e m e n t b e t w e e n r e m o t e a n d onsite sur- T a b l e I I I . W o u n d questionnaire* Wound descriptors Present Gangrene Necrosis Erythema CeHufitis/infection Iscbemiat Granulation tissue Wound management decisions Yes Wound healing problem present Need for examination within 24 hr by MD Need for hospitalization Need for antibiotics Need for debridement Not present No *The following additional wound descriptors and management decisions were asked but not used for analysis because of either infrequent occurrence, limited clhlical significance, vagueness of surgeon response, or poorly phrased question: (1) ecchymosis; (2) exposed bypass graft; (3) nonhealing wounds; (4) exposed tendon/bone; (5) edema; (6) drainage; (7) bedrest; (8) dressing changes; and (9) leg elevation. tThis descriptor was added to the questionnaire 4 weeks into the study. g e o n s was c a l c u l a t e d in t h r e e d i f f e r e n t subsets o f cases based o n the level o f a g r e e m e n t a m o n g onsitc surgeons: (1) all cases; (2) o n l y those cases in which onsite s u r g e o n a g r e e m e n t was greater t h a n 67%; a n d (3) only those cases in which onsite a g r e e m e n t was 100%. Ultimately, a g r e e m e n t was m e a s u r e d in five different categories: c a t e g o r y I , a g r e e m e n t a m o n g o n s i t e s u r g e o n s ; c a t e g o r y I I , a g r e e m e n t a m o n g r e m o t e surgeons; c a t e g o r y I I I , a g r e e m e n t b e t w e e n r e m o t e a n d onsite surgeons for all cases; c a t e g o r y IV, a g r e e m e n t b e t w e e n r e m o t e and onsite surgeons only in t h o s e cases in w h i c h onsite surgeons a g r e e m e n t was g r e a t e r t h a n 67%; a n d c a t e g o r y V, a g r e e m e n t bct~vccn r e m o t e a n d onsite s u r g e o n s o n l y in t h o s e cases in which onsite surgeons a g r e e m e n t was 100%. T o d e t e r m i n e f a c t o r s t h a t i n f l u e n c e d c o n c o r - dance, logistic regressions with d e p e n d e n t variables (certainty o f response, a g r e e m e n t b e t w e e n onsite a n d r e m o t e surgeons, a n d d i s a g r e e m e n t b e t w e e n onsite a n d r e m o t e s u r g e o n s ) a n d e x p l a n a t o r y variables (onsite vs r e m o t e surgeon, level o f surgical training, a n d w o u n d type) w e r e p e r f o r m e d . D a t a sets for m u l - tivariate analysis were g e n e r a t e d using the SAS-based r a n d o m n u m b e r g e n e r a t o r t o r a n d o m l y e l i m i n a t e r e m o t e a n d onsite evaluations in excess o f t w o per w o u n d . A p value o f 0.01 was considered significant r a t h e r t h a n using a p value o f 0.05 w i t h B o n f e r r o n i c o r r e c t i o n for multiple comparisons. Sensitivity a n d specificity o f r e m o t e w o u n d eval- u a t i o n were calculated f o r w o u n d diagnoses, includ- ing p r e s e n c e o f a w o u n d healing p r o b l e m , necrosis, JOURNAL OF VASCULAR SURGEP.Y Volume 27, Number 6 W i r t h l i n et al. 1 0 9 3 Fig. 2. Example of images representing each wound category. A, Postoperative incision; B, amputanon; C, gangrenous/necrotic toes; D, nonhealing ulcer. Wounds were photographed from vascular surgery inpatients with a Kodak DC50 digital camera during morning rounds. ischemia, erythema, granulation, gangrene, and cel- lulitis/infection. Results were derived from two sub- sets o f cases--greater than 67% agreement among onsite surgeons (category IV) and 100% agreement among onsite surgeons (category V). Sensitivity was calculated as the number o f "present" responses by remote surgeons divided by the total number o f opportunities for remote surgeons to respond "pre- sent" when onsite surgeons agreed (as defined in category IV and V) to "present." Specificity was cal- culated as the number " n o t present" responses by remote surgeons divided by the total number o f opportunities for remote surgeons to respond "pre- sent" when onsite surgeons agreed (as defined in category IV and V) to " n o t present." R E S U L T S There were no wound complications associated with the imaging process, and imaging took approx- imately 5 to 7 minutes per patient. Remote surgeons spent an average o f 3 to 5 minutes viewing images o f each patient. The mean quality index for all image sets was 3.3, with a range o f 2.7 to 3.8, and none o f the images were considered unusable. For w o u n d descriptors, the average certainty ranged from 9.5 to 9.9 among onsite surgeons and 8.2 to 9.8 among remote surgeons. Certainty was lowest for detection o f ischemia and cellulitis among both onsite and remote surgeons. For wound man- agement decisions, the average certainty ranged from 9.6 to 9.8 among onsite surgeons and 9.4 to 9.7 among remote surgeons. Certainty was lowest for use o f antibiotics and need for hospitalization among both onsite and remote surgeons. Attending surgeons compared with other evaluators recorded lower cer- tainty (certainty less than 10) when diagnosing ery- thema (odds ratio [OR], 0.06; p = 0.007). Otherwise, certainty was not influenced by remote versus onsite location, level of training, or wound type. Table IV shows the average percent agreement and kappa values regarding wound descriptors, and Table V shows the average percent agreement and kappa values regarding wound management deci- sions. Prevalence data, kappa, kappa(nor), and kappa(max) for category V are shown in Table VI. The specificity and sensitivity o f remote wound diag- nosis are shown in Table VII. The level o f surgical training was not associated (p > 0.01) with agreement or disagreement between remote and onsite surgeons for both wound descrip- tion and management. Wound type (nonhealing ulcer compared with postoperative incision) posi- tively influenced concordance between remote and onsite surgeons when diagnosing gangrene (OR, J O U R N A L OF VASCULAR SURGERY 1 0 9 4 W i r t h l i n et aL June 1998 Table IV, Average percent agreement regarding w o u n d descriptors Wound descriptors Category I Category H * C a t e g o r y I I I tCategory I V ¢Category V Agreement among Agreement among Agreement b e t w e e n Agreement b e t w e e n Agreement between onsite surgeons remote surgeons onsite surgeons vs onsite surgeons vs onsite surgeons vs (n = 45) (n = 45) remote surgeons (n = 45) r e m o t e surgeons remote surgeons N e c r o s i s G r a n u l a t i o n t i s s u e I s c h e m i a § G a n g r e n e C e l l u l i t i s / i n f e c t i o n E r y t h e m a 80% K = 0 . 5 3 9 0 % ~ = 0 . 3 3 83% K = 0 . 6 0 81% K = 0 . 3 4 9 2 % K = 0 . 4 4 80% K = 0 . 4 4 80% lc = 0 . 7 0 83% ~ = 0 . 5 0 78% K = 0 . 7 0 85% ~c = 0 . 7 7 9 3 % K = 0 . 6 4 74% K = 0 . 5 5 67% K = - 0 . 0 4 81% )c = 0 . 0 8 62% lc = 0 . 2 4 64% ~c = 0 . 2 2 86% K = 0 . 2 8 60% ~; = 0 . 1 2 91% ( n = 3 7 ) I< = 0 . 7 8 91% ( n = 3 3 ) ~: = 0 . 6 4 91% ( n = 8 ) ~ = 0 . 0 0 77% ( n = 3 9 ) K = 0 . 5 3 65% ( n = 3 3 ) ~ = 0 . 0 8 65% ( n = 3 0 ) ~ = 0 . 0 2 95% ( n = 3 4 ) ~ = 0 . 9 0 93% ( n = 3 1 ) ~ = 0 . 4 5 91% ( n = 7 ) ~ = 0 . 0 0 81% ( n = 3 7 ) ~ = 0 . 6 2 69% ( n = 2 7 ) )c = 0 . 0 0 66% ( n = 2 8 ) 1( = 0 . 0 2 * C o m p a r i s o n m a d e in all cases. t C o m p a r i s o n m a d e o n l y in s u b s e t o f cases w h e n 9 ; C o m p a r i s o n m a d e o n l y in s u b s e t o f cases w h e n § I s c h e m i a d a t a a v a i l a b l e f o r 1 0 p a t i e n t s only. ~c = k a p p a v a l u e . o n s i t e s u r g e o n s a g r e e d g r e a t e r t h a n 67%. o n s i t e s u r g e o n s a g r e e d 100%. Table V. Average percent agreement regarding w o u n d management decisions Wound descriptors Category I Category H * C a t e g o r y I I I /'Category I V ~Category V Agreement among Agreement among Agreement b e t w e e n Agreement b e t w e e n Agreement between onsite surgeons r e m o t e surgeons onsite surgeons vs onsite surgeons vs onsite surgeons vs (n = 45) (n = 45) r e m o t e surgeons (n = 45) r e m o t e surgeons remote surgeons W o u n d h e a l i n g 95% )c = 0 . 7 5 95% )c = 0 . 3 1 87% K = 0 . 4 3 92% ( n = 3 9 ) ~ = 0 . 4 8 p r o b l e m E m e r g e n t e x a m - 64% ~ = 0 . 1 7 80% ~ = 0 . 4 1 74% )c = 0 . 5 9 89% ( n = 2 9 ) ~ = 0 . 4 1 i n a t i o n N e e d f o r h o s p i t a l - 85% K = 0 . 0 4 85% ~: = 0 . 1 5 79% ~: = 0 . 2 6 87% ( n = 4 2 ) ~z = 0 . 6 7 i z a t i o n A n t i b i o t i c s 68% K = 0 . 2 8 80% ~ = 0 . 0 5 63% ~ = 0 . 3 9 71% ( n = 3 8 ) ~ = 0 . 3 9 D e b r i d e m e n t 74% K = 0 . 5 0 78% ~ = 0 . 1 8 60% K = 0 . 3 0 66% ( n = 3 8 ) ~ = 0 . 3 0 91% ( n = 3 9 ) ~: = 0 . 4 8 86% ( n = 2 6 ) ~ = 0 . 4 4 84% ( n = 3 9 ) lc = 0 . 8 0 69% ( n = 3 0 ) ~c = 0 . 4 0 63% ( n = 3 5 ) ~ = 0 . 2 3 * C o m p a r i s o n m a d e i n all cases. t C o m p a r i s o n m a d e o n l y in s u b s e t o f cases w h e n o n s i t e s u r g e o n s a g r e e d g r e a t e r t h a n 67%. : ~ C o m p a r i s o n m a d e o n l y i n s u b s e t o f cases w h e n o n s i t e s u r g e o n s a g r e e d 100%. ~: = k a p p a v a l u e . 4.1; p = 0 . 0 0 0 1 ) , granulation ( O R , 17.4; p = 0.0001), and determining the need for debridement (OR, 10.0; p = 0.0001). However, nonhealing ulcer was associated with decreased agreement between remote and onsite surgeons regarding n e e d for antibiotics (OR, 0.4; p = 0.009) and need for hospi- talization (OR, 0.2; p = 0.0003) and increased dis- agreement b e t w e e n remote and onsite surgeons regarding cellulitis/infection (OR, 3.1; p = 0.006) and need for emergent examination (OR, 8.6; p = 0.0001). D I S C U S S I O N Telemedicine has been in development for more than 30 years and has experienced rapid growth in the 1990s. Proponents o f telemedicine envision sev- eral valuable applications: providing specialty care to underserved areas, increasing the efficiency o f exist- ing medical resources, expanding a hospital's service area, and attracting international health care dollars to the U n i t e d States. Moreover, Perednia and Allen 18 predict that by 2000 many physicians will be directly or indirectly involved in clinical telemedi- cine. Telemedicine is already an integral tool in most radiology practices, and thcrc is ongoing develop- m e n t and clinical investigation o f telemedicine in almost all medical fields. This investigation is the first to critically evaluate the feasibility o f using digital images for remote w o u n d management. The main objective o f this study was to establish the " p r o o f o f concept" o f dig- ital imaging before implementing this technology in vascular surgery h o m e care. Implementation o f telemedicine requires validation o f quality o f care, ease o f use, cost-effectiveness, and acceptance by both patients and physicians. Our observations from this preliminary investigation suggest that digital photography in conjunction with telemedicine can IOURNAL OF VASCULAR SURGERY Volume 27, Number 6 Wirthlin et aL 1 0 9 5 Table VI. Kappa and prevalence data for wound management decisions and descriptors in category V* Kappa(nor)t Kappa(max)t Kappa ( 2 P o - l ) (po2/[1-Po]2+1) Atrue(+) Blaise(-) Cfalse(+) Dtrue(-) n~ W o u n d m a n a g e m e n t decisions Wound healing problem 0.48 0.79 0.79 55 2 5 4 33 Emergent examination 0.44 0.44 0.48 20 4 10 16 25 Need for hospitalization 0.80 0.83 0.83 13 3 1 29 23 Antibiotics 0.40 0.40 0.45 19 7 11 23 30 Debridement 0.23 0.70 0.70 2 i 9 54 33 W o u n d descriptors Necrosis 0.90 0.91 0.91 43 1 2 20 33 Granulation tissue 0.45 0.67 0,68 6 2 8 44 30 Ischemia§ 0.00 0.67 0.68 5 1 0 0 3 Gangrene 0.62 0.62 0.64 25 7 7 35 37 Cellulifis/infection 0.00 0.22 0.32 0 0 21 33 27 Erythema 0.02 0.19 0.30 28 6 16 4 27 *Comparison made only in subset o f cases when onsite tPo = proportion o f agreements as defined by Lantz et l:n = number o f cases. §Ischemia data available for lO patients only. surgeons agreed 100%. al.17 See appendix for explanation o f Kappa(nor) and Kappa(max). Table VII. Sensitivity and specificity for wound descriptors *Agreement category IV ?Agreement category V Descriptor Sensitivity Specificity Sensitivity Specificity Wound healing problems 98% 53% 98% 53% Necrosis 98% 82% 98% 87% Ischemia 88% 100% 88% 100% Erythema 87% 26% 89% 27% Granulation 77% 97% 82% 96% Gangrene 75% 82% 78% 85% Cellulitis/infection 71% 65% NA 66% *Calculations performed on subset o f cases in which onsite surgeons agreed greater than 67%. tCalculations performed on subset o f cases in which onsite surgeons agreed 100%. provide quality remote wound care using a relative- ly simple and cost-effective protocol. The potential quality o f remote w o u n d care was determined using bedside examination as the frame o f reference and evaluating concordance between r e m o t e and onsite surgeons. C o n c o r d a n c e o f response based on viewing various imaging media versus more conventional evaluation (hard-copy radiographs, glass slides, or patient examination) has been used to validate new technology in telera- diology, telepathology, and teledermatology.9,19-21 In our study, agreement between remote and onsite surgeons for b o t h w o u n d diagnosis and management was high (63% to 95%) and was com- parable with concordance observed in telederma- tology (83%). 9 Using onsite evaluation as a refer- ence, remote surgeons were able to make equiva- lent diagnoses in 66% to 95% o f image sets and r e c o m m e n d comparable m a n a g e m e n t in 66% to 92% o f cases. We also observed that the level o f dis- agreement between remote and onsite surgeons equaled the disagreement among onsite surgeons, suggesting that the imaging media did not inde- p e n d e n t l y impact agreement. Moreover, w o u n d healing problems and specific w o u n d conditions are readily detected by digital images, as demon- strated by the sensitivity o f remote wound diagno- sis (71% to 98%). These observations suggest that in general remote wound management via digital imaging will render care comparable with conven- tional evaluation. Kappa values generated trends similar to percent agreement, that is, the range of kappa values among onsite surgeons (-0.04 to 0.75) was comparable with that o f remote versus onsite surgeons (0.00 to 0.90), and when kappa values were low among onsite sur- geons, values were similarly low between remote and onsite surgeons. I n only one parameter (need for JOURNAL OF VASCULAR SURGERY 1 0 9 6 W i r t h l i n e t al. June 1998 debridement) was the kappa value in category V less than 0.4, whereas thc corrcsponding kappa valuc in category I was greater than 0.4 (a kappa value less than 0.4 denotes "marginal reproducibility"). 23 Our data also depicted the base rate problcm in kappa sta- tistics described by Lantz et al. 17 and Spitznagel et al.22 in which kappa values may bc skewed by preva- lence. For example, in several instances (i.e., presence o f w o u n d healing problem, granulation tissue, and ischemia) high percent agreement was associated with a low kappa value because o f unevenly distrib- uted prevalence. Correcting for uneven prevalence using kappa(nor) increased the concordance between onsite and remote surgeons (see appendix). Thus when there was reproducible agreement among onsite surgeons, digital imaging succeeded in that agreement between remote and onsite surgeons was high in these cases. Certainty o f remote w o u n d evaluation and man- agement was no different than that o f onsite sur- geons, and contrary to the experience in telederma- tology the certainty o f diagnosis varied little and did n o t influence agreement. 9 However, the certainty level was universally high, which may represent a response construct among surgeons (surgeons are more likely to respond " 1 0 " - - c o m p l e t e l y certain). Nonetheless, viewing digital imagcs did n o t appear to influence surgeons' certainty o f w o u n d diagnosis and management. Using our imaging protocol, remote w o u n d care may be limited by decreased ability to accurately rep- resent erythema. Agreement between remote and onsite surgeons was lowest for erythema (66%), cel- lulitis (69%), and management decisions that rou- tinely follow the diagnosis o f cellulitis, such as antibiotics (66%) and urgent examination (71%). Several factors may explain this observation includ- ing variable sensitivity o f surgeon examination, learning curve o f remote diagnosis, and purely tech- nological factors related to digital imaging. O f these, variation o f surgeon examination appeared to have the greatest impact, as shown by percent agreement and kappa values (poor onsite agreement appeared to be associated with decreased agreement between remote and onsite surgeons). Variation among physicians has been documented in other fields 24-31 and has been cited as a limitation o f evaluating w o u n d infections after vascular surgery, ll Thus without a criterion standard for w o u n d descrip- tion or management, we used bedside examination as the "gold standard" and observed considerable dis- agreement among onsite surgeons. For example, onsite surgeons disagreed in approximately one third o f cases regarding the presence o f erythema or cel- lulitis and regarding important management deci- sions such as need for urgent evaluation and hospi- talization. Also, we observed similar variation among remote surgeons. Thus physician variability, as in other investigations, limited our ability to evaluate this technology and may have adversely impacted concordance regarding erythema and cellulitis. In an attempt to control for surgeon variation, we stratified cases according to level o f onsite agreement and observed increasing concordance between remote and onsite surgeons (cellulitis, 62% to 65% to 69%). Unfortunately, because o f small patient numbers we were unable to simultaneously control for remote and onsite surgeon variability. Finally, physician vari- ation observed in this study and other investigations o f w o u n d complications speaks to the complexity o f managing vascular w o u n d s and begs for evalua- tion/technology that can more objectively evaluate erythema and cellulitis. There is likely a learning curve related to remote w o u n d management that may have impacted concor- dance o f diagnosing crythema and ccllulitis. For example, multivariate analysis showed that remote surgeons were less certain o f diagnosing cellulitis compared with making other diagnoses. Multivariate analysis also showed that disagreemcnt rcgarding erythema and cellulitis was more likely for nonheal- ing ulcers comparcd with other w o u n d types. We observed that remote surgeons appeared more hesi- tant to render diagnoses and treatments during their initial remote evaluations compared with later evalu- ations. Moreover, remote surgeons "overdiagnosed" and "overtreated" wounds compared with onsite sur- geons, that is, remote surgeons more often diag- nosed erythema and more often prescribed antibi- otics. This is depicted in the high number o f false positive results when remote surgeons disagreed with onsite surgeons (16 o f 22 for erythema and 21 o f 21 for cellulitis; Table VI). Accordingly, the sensitivity o f remotely diagnosing erythema was high (87%), whereas the specificity was low (26%). Thus, at pre- sent, erythema is readily diagnosed and may at times be "overdiagnosed" because o f extra caution on the part o f remote surgeons. However, with continued experience we would expect the accuracy o f detecting and treating erythema/cellulitis to improve. Technological factors may have also had an impact on the accuracy o f diagnosing erythema and cellnlitis. Numerous factors influence image quality, including the photographer, lighting, resolution, and computer monitor. We purposely chose a digital camera priced at less than $1000 to test a protocol JOURNAL OF VASCULAR SURGERY Volume 27, Number 6 W i r t h l i n et ai. 1097 t h a t w o u l d be feasible from a cost perspective w h e n i m p l e m e n t e d in nursing h o m e care. The resolution o f the camera used in this s t u d y (756 x 506 pix- e l s / i n 2) has been shown to be adequate compared with higher resolutions for clinical diagnosis in tele- d e r m a t o l o g y , a2 T h e single color chip, however, appears to be very sensitive to lighting and contains an infrared filter that depicts ultraviolet light as red color. This may have overrepresented red tones in images o f vascular wounds. Subsequently, we have observed that lighting (incandescent versus fluores- cent) alters the r e p r e s e n t a t i o n o f e r y t h e m a in patients with lower extremity cellulitis (unpublished data) and have since used polarizing filters to correct this characteristic o f singlechip digital cameras. Despite this potential flaw, the image quality is excel- l e n t and provides e n o u g h visual i n f o r m a t i o n t o appropriately m a n a g e the m a j o r i t y o f w o u n d s . Moreover, there are several cameras with higher res- olution and multiple color sensors t h a t m a y be need- ed to optimize remote w o u n d management. Finally, this t e c h n o l o g y omits i m p o r t a n t c o m p o n e n t s o f physical examination, such as palpation and olfac- tion, t h a t m a y allow for m o r e accurate diagnosis. N o n e t h e l e s s , digital i m a g i n g for r e m o t e w o u n d diagnosis is i n t e n d e d as an adjunct to standard phys- ical e x a m i n a t i o n p e r f o r m e d by the local care provider ( h o m e care nurse or physician). O u r observations s u g g e s t t h a t h e a l t h care providers o f varied levels o f training and background can successfully i m p l e m e n t this t e c h n o l o g y with rel- atively low costs in terms o f b o t h time and equip- ment. All images were taken by an individual with no medical or photographic background after mini- mal training (6 to 8 hours). The photographic qual- ity was universally g o o d , and all images were consid- ered clinically useful. Moreover, the imaging process was quick (5 to 7 minutes per patient) and did n o t complicate patient care. Since this project, vascular h o m e care nurses have been trained in a short peri- od o f time (about 2 weeks) to p h o t o g r a p h w o u n d s a n d t r a n s m i t images f r o m patients' h o m e s to the attending surgeon (Fig. 3). A l t h o u g h we did n o t specifically evaluate cost, several observations suggest t h a t this telemedicine application will be cost-effective. First, the digital camera is relatively inexpensive a n d decreased in price d u r i n g the study period ($800 to $500). We expect t h a t the cost o f this t e c h n o l o g y will continue to decrease. Second, the e q u i p m e n t to view images is simple and relatively inexpensive (standard desk top computer, monitor, and software), and current- ly attending surgeons can access images o f patients' Fig. 3. Wound images transmitted from a patient's home to the attending vascular surgeon during a home care visit. Images were taken by home-care nurses and transmitted with a lap top computer and standard phone line. A, Time zero. B, One month later. w o u n d s from their office computers or any o t h e r c o m p u t e r connected to the Massachusetts General Hospital network. The greatest cost o f implement- ing o f this telemedicine application has been the lap- top computers ($3500) used by h o m e care nurses to t r a n s m i t images f r o m p a t i e n t s ' h o m e s . This cost could be easily offset by fewer office visits, emer- gency r o o m evaluations, and shortened hospital stay. However, the ultimate cost-effectiveness will need to be proven in a prospective trial. Finally, o u r experi- ence suggests that physician and patient acceptance o f this t e c h n o l o g y is high, and currently nearly all eligible o u t p a t i e n t s c o n s e n t to digital i m a g i n g o f their wounds. However, a prospective trial o f r e m o t e w o u n d m a n a g e m e n t in outpatients, evaluating the opera- tional feasibility, quality o f care, and cost-effective- ness is n e e d e d to f u r t h e r validate this technology. Mso, further research o f remote w o u n d care m u s t be coupled with accurate assessment o f cost and quality JOURNAL OF VASCULAR SURGERY 1 0 9 8 W i r t h l i n et aL June 1998 o f outpatient care, which at present, despite much data o n inpatient care, are u n l ~ n o w n . 33 Moreover, recent reports suggest that quality o f home care may affect medical costs and outcomes, 33-35 which cor- roborates our experience with specialized vascular h o m e care. Implementation o f remote w o u n d sur- veillance along with other applications in nursing home care could improve the quality o f outpatient care and provide specialized care to remote locations. Also, we did not investigate the potential o f image manipulation such as simultaneous display o f wounds at varying time intervals, w o u n d area measurement, and outlining o f specific shades o f redness. These techniques could allow for more objective w o u n d evaluation. Finally, issues related to confidentiality, licensure, liability, and reimbursement, which have not been entirely resolved in other fields o f telemed- icine, must be considered in conjunction with devel- opment o f this technology. 18 Certainly, technology in digital imaging and elec- tronic transfer o f data will advance independent o f health care trends, and continued research in telemed- icine should capitalize on this progress. Technologic advancements that would complement digital imag- ing for remote w o u n d management include: (1) development o f an electronic patient record that incorporates digital images, radiographs, noninvasive laboratory data, and the hospital chart; and (2) devel- opment o f equipment to collect and transmit vascular noninvasive data for remote graft surveillance. C O N C L U S I O N This study suggests that digital imaging for remote w o u n d management is feasible on the basis o f high concordance between remote and onsite surgeons regarding w o u n d evaluation and management. This application o f telemedicine has the potential to improve the quality o f current outpatient wound care while decreasing the cost by allowing home care nurs- es to photograph wounds and transmit images over any distance in a short period o f time. This hypothesis needs to be verified in a prospective clinical trial assess- ing clinical outcomes and medical costs. The available technology in image media and electronic data trans- fer is advanced; however, the medical application o f this technology is in its infancy. Telemedicine holds significant promise in vascular surgery but will require careful, well-designed research for development and validation o f its clinical usefulness. We thank the Eastman Kodak Co. f o r supplying equip- m e n t (Kodak D C 5 0 camera), and w e gratefully acknowl- edge t h e assistance o f t h e n o n a u t h o r physicians w h o evalu- ated w o u n d s for this study, specifically Jeffrey Slaiby, M D , Peter Purcell M D , Joseph Giglia, M D , and a n u m b e r o f General S u r g e r y Residents at Massachusetts General Hospital. I n a d d i t i o n , w e a c k n o w l e d g e t h e assistance o f Virginia Capasso, RN, Debbie Burke, RN, and t h e other nurses on Bigelow 14 (Massachusetts General Hospital), Eric R. M e n n , Kimberly D. Galbraith, Linda A. M o t t l e , and o t h e r T e l e m e d i c i n e Center personnel, and Philip A A m a t o , PhD, w h o s u p p l i e d t h e Kodak D C 5 0 for o u r use. R E F E R E N C E S 1. Bowersox JC, Shah A, Jensen J, Hill H, Cordts PR, Green PS. Vascular applications o f telepresence surgery: initial feasi- bility studies in swine. J Vasc Surg 1996;23:281-7. 2. Grigsby J, Kaehny MM, Sandbert EJ, Schlenker RE, Shanghnessy PW. Effects and effectiveness o f telemedicine. Health Care Financing Rev 1995;17:27-34. 3. Perednia DA, Allen A. Telemedicine technology and clinical applications. JAMA 1995;273:483-8. 4. Hassol A, Ganmer G, Grigsby J, Mintzer CL, Puskin DS, Brunswick M. Rural telemedicine: a national snapshot. Telemed J 1996;2:43-8. 5. Federman D, Hogan D, Taylor JR, Caralis P, Kirsner RS. A comparison of diagnosis, evaluation, and treatment o f patients with dermatologic disorders. J Am Acad Dermatol 1995;32:726-9. 6. Herman PG, Gerson DE, Hessel SJ, Mayer BS, Wamick M, Blesser B, et al. Disagreements in chest roentgen interpreta- tion. Chest 1975;63:278-82. 7. Perednia DA, Brown NA. Teledermatology: one application o f telemedicine. Bull Med Libr Assoc 1995:83(1):42-7. 8. Menn ER, Kvedar JC. Teledermatology in a changing health care environment. Telemedicine 1995;1:303-8. 9. Kvedar JC, Edwards RA, Menn ER, Maofid M, Gonzalez E, Dover J, et ai. The substitution o f digital images for derma- tologic physical examination. Arch Dermatol. 1997;133: 161-7. 10. Johnson JA, Cogbill TH, Strutt PJ, Gundersen AL. Wound complications after infrainguinal bypass. Arch Surg 1988; 123:859-62. 11. Reifsnyder T, Bandyk D, Seabrook G, Kinney E, Towne J. Wound complications o f the in situ saphenous vein bypass technique. J Vasc Surg 1992;15:843-50. 12. Donaldson MC, Mannick JA, Whittemore AD. Femoral-dis- tal bypass with in situ greater saphenous vein. Ann Surg 1991;213:457-65. 13. Szilagyi DE, Smith RF, Elliott JP, Vrandecie MP. Infection in arterial reconstruction with synthetic grafts. Ann Surg 1972;176:321-33. 14. Calligaro KD, Dougherty MJ, Raviola CA, Musser DJ, DeLanrentis DA. Impact o f clinical pathways on hospital costs and early outcome after major vascular surgery. J Vasc Surg 1995;22:649-57. 15. Patterson RB, Whitley D, Porter K. Critical pathways and cost-effective practice. Semin Vase Surg 1997;10(2):113-8. 16. PC Webopaedia (online). Available from: www.pcwebope- d i a . c o m / J P E G . h t m / ( 1 9 9 7 , Sep. 2). 17. Lantz CA, Nebenzahl E. Behavior and interpretation o f the K statistic: resolution o f the two paradoxes.-J Clin Epidemiol 1996;49:431-4. 18. Perednia DA, Allen A. Telemedicine technology and clinical applications. JAMA 1995;273:483-8. JOURNAL OF VASCULAR SURGERY Volume 27, Number 6 Wirthlin et al. 1 0 9 9 19. Zelickson BD, Homan L. Teledermatology in the nursing home. Arch Dermatol 1997;133:171-4. 20. Halliday BE, Bhattacharyya AK, Graham AR, Davis JR, Leavitt SA, Nagle RB, et al. Diagnostic accuracy of an inter- national static-imaging telepathology consultation service. Hum Pathol 1997;28(1):17-21. 21. Scott WW, Rosenbaum JE, Ackerman SJ, Reichle RL, Magid D, Weller JC, et al. Subtle orthopedic fractures: teleradiolo- gy workstation versus film interpretation. Radiology 1993; 187:811-5. 22. Spitznagel EL, Helzer IE. A proposed solution to base rate problem in the kappa statistics. Arch Gcn Psychiatry 1985; 42:725-8. 23. Rosner B. Hypothesis testing: categorical data. In: Kugushev A, editor. Fundamentals of biostatistics. Belmont, Calif.: Wadworth Publishing; 1995. p. 426. 24. Garland LH. Studies on the accuracy of diagnostic proce- dures. AJR Am J Roentgenol 1959;82:25-38. 25. Hillman BJ, Hessel SJ, Swensson RC, Herman PG. Improving diagnostic accuracy: a comparison of interactive delphi consultations. Invest Radiol 1977;12:112-5. 26. Diagnostic decision-process in suspected pulmonary embolism: report oft_he Herlev Hospital study group. Lancet 1979;1:1336-8. 27. Bader JD, Shugars DA. Variation in dentists' clinical deci- sions. J Public Health Dent 1995;55:181-8. 28. Davis PB, gee RL, Millar J. Accounting for medical variation: the case of prescribing activity in a New Zealand general prac- tice sample. Soc Sci Med 1994;39:367-74. 29. Records NL, Tomblin JB. Clinical decision making: describ- ing the decision rules of practicing speech-language patholo- gists. J Speech Lang Hear Res 1994;37:144-56. 30. Leunens G, Menten J, Weltens C, Verstraete J, van der Schueren E. Quality assessment of medical decision making in radiation oncology: variability in target volume delineation for brain tumours. Radiother Oncol 1993;29:169-75. 31. Lidegaard O, Bottcher LM, Weber T. Description, evaluation and clinical decision making according to various fetal heart rate patterns: inter-observer and regional variability. Acta Obstet Gynecol Scand 1992;71:48-53. 32. Bittorf A, Fartasch M, Schuler G, Diepgen TL. Resolution requirements for digital images in dermatology. I Am Acad Dermatol 1997;37:195-8. 33. Dahlberg NL. A perinatal center antepartal homecare pro- gram. l Obstet Gynccol Neonatal Nuts 1988;17:30-4. 34. Donleavy J. Responsive restructuring: part I. Acute care nursing provider home visits. New Definition 1993;8:1-3. 35. Heaman M, Thompson L, Helewa M. Patient satisfaction with an antepartal home care program. J Obstet Gynecol Neonatal Nurs 1994;23:707-13. Submitted Sep. 22, 1997; accepted Feb. 11, 1998. A P P E N D I X . A s s h o w n b y L a n t z e t al. 17 a n d S p i t z n a g e l e t al. 22 k a p p a v a l u e s c a n v a r y f o r a g i v e n l e v e l o f a g r e e m e n t d e p e n d i n g o n t h e p r e v a l e n c e o f a g r e e m e n t s ( t r u e p o s i t i v e s a n d t r u e n e g a t i v e s ) a n d d i s a g r e e m e n t s (false p o s i t i v e s a n d false n e g a t i v e s ) . I L a p p a ( n o r ) , a p a r a m e - t e r d e r i v e d f r o m p r o p o r t i o n o f a g r e e m e n t s ( P o ) , has b e e n s u g g e s t e d as a s o l u t i o n f o r t h e v a r i a b i l i t y o b s e r v e d i n k a p p a v a l u e s w h e r e P o = (a + d ) / N a n d K a p p a ( n o r ) = 2 P o - 1. K a p p a ( n o r ) c o r r e c t s f o r a s y m m e t r y o f p r e v a l e n c e a n d is e q u a l t o k a p p a o n l y w h e n t h e r e is s y m m e t r y o f b o t h a g r e e m e n t s a n d dis- a g r e e m e n t s . K a p p a ( m a x ) is t h e m a x i m u m k a p p a v a l u e f o r a g i v e n p r e v a l e n c e o f a g r e e m e n t a n d d i s - a g r e e m e n t . K a p p a ( m a x ) b a l a n c e s a g r e e m e n t i n t h e t w o a g r e e m e n t c a t e g o r i e s a n d m a x i m a l l y s k e w s d i s - a g r e e m e n t i n t h e d i s a g r e e m e n t c a t e g o r i e s u s i n g t h e equation. Kappa(max) = P o 2 / (1 - P o ) 2 + 1. S e e ref- e r e n c e s 1 7 a n d 2 2 f o r f u r t h e r d i s c u s s i o n . D I S C U S S I O N Dr. J a m e s E s t e s (Boston, Mass.). I applaud y o u r inter- est in l o o k i n g in the technologic front here in terms o f c o m b i n i n g computers and medicine, and I enjoyed y o u r talk. I have o n e q u e s t i o n f r o m a p r a c t i c a l s t a n d p o i n t : where do you see this t e c h n o l o g y being applied, specifi- cally, in terms o f justifying the costs o f the e q u i p m e n t for acquiring and transmitting digital information? Dr. D o u g l a s J. W i r t h l i n . I think the potential for this t e c h n o l o g y is e n o r m o u s . We e m b a r k e d o n this p r o j e c t planning to i m p l e m e n t digital imaging in o u t p a t i e n t man- agement o f vascular wounds to decrease cost o f care while maintaining quality o f care. The cost o f e q u i p m e n t is min- imal c o m p a r e d With the potential cost savings. However, this needs t o be proven in a prospective trial. Dr. C a r l E. B r e d e n b e r g (Portland, Me.). O u r experi- ence thus far at Maine Medical Center with these types o f techniques has been limited largely to shared educational conferences, particularly in vascular surgery, with Dave Pilcher and Michael Ritchie at the University o f Vermont, and it's remarkable I think h o w lively and alive this is w h e n you are using n o t these still cameras b u t with television. I ' m still n o t sure, and this is in p a r t perhaps in answer to the previous question, whether this is the g o o d news or the bad news from an i n s t i t u t i o n ' s p o i n t o f view. F o r a rural network, for example, I could argue t h a t to be able to visualize wounds o f a patient up in Caribou, to visual- ize those d o w n in P o r t l a n d , and make decisions a b o u t whether o r n o t to get the patient to P o r t l a n d is clearly the g o o d news. N o w if, however, i t is a p a t i e n t in S o u t h P o r t l a n d and this is being viewed on F r u i t Street at the Massachusetts General Hospital, then I ' m less persuaded that this is truly g o o d news. This is n o t entirely specula- tion. F o r example, at Mercy Hospital in P o r t l a n d the radi- o l o g y d e p a r t m e n t s t a t i o n e r y carries t h e l o g o o f the JOURNAL OF VASCULAR SURGERY 1 1 0 0 W i r t h l i n et al. June 1998 M a s s a c h u s e t t s G e n e r a l H o s p i t a l r a d i o l o g y d e p a r t m e n t along with their own, so the implications o f this technol- o g y are i n d e e d far-reaching in many ways. Dr. W i r t h l i n . I n regards t o y o u r first c o m m e n t , y o u m e n t i o n e d that the video imaging was very useful. We use still digital images because the infrastructure requirements for t r a n s m i t t i n g v i d e o messages are m u c h g r e a t e r a n d m o r e expensive. Also, the resolution o f a still digital image is better than t h a t o f a video image. I n regards to whether this is g o o d news or bad news, I think from a patient s t a n d p o i n t it is g o o d news as l o n g as the t e c h n o l o g y is d e v e l o p e d and i m p l e m e n t e d properly. I n terms o f h o w this t e c h n o l o g y may be used at m a j o r medical centers is unclear at this time, and really this tech- n o l o g y is in its infancy. T h e r e is m u c h w o r k t h a t needs to be d o n e before this can be i m p l e m e n t e d , for example, a clinical trial proving t h a t this is safe and cost-effective. Dr. D a v i d B. P i l c h e r (Colchester, Vt.). This is a very disturbing paper, because it seems to me t h a t y o u are say- ing t h a t we surgeons w h o talk t o patients, who interact with patients w h e n we are visualizing their wounds, can n o w be radiologists and just l o o k at the still pictures. I d o n ' t t h i n k y o u r conclusion t h a t telemedicine has a value is n o t really valid. Your conclusion is that looking at still pictures is o f value. T h a t ' s n o t really telemedicine. T h a t ' s a l o t cheaper than telemedlcine, as I think you just s a i d - - transmitting digital pictures d o e s n ' t require fifll telemedi- cine, n o r does it allow the full potential, so I d o n ' t think y o u are really l o o k i n g at telemedicine. Dr. W i r t h l i n . T h e definition o f t e l e m e d i c i n e is v e r y broad. Telemedicine can be as simple as a p h o n e consulta- tion, a 911 p h o n e call, o r can be as complex as teleprcsence surgery. This project was developed just to test the feasibil- ity o f digital imaging. I n o t h e r words, h o w accurately does a digital image represent a wound? This is just the first step in the development o f this technology, and y o u ' r e tight, we did n o t test the telemedicine application o f digital imaging yet. T h a t is the next project, in which h o m e care nurses will transmit the images to the treating surgeons. T h e process o f transmitting the digital images makes it a telemedicine application. You're right, viewing a digital image is n o t equal to a standard physical examination, b u t it is comparable. Also, in terms o f the impact this t e c h n o l o g y may have o n the h u m a n e l e m e n t o r p a t i e n t - p h y s i c i a n r e l a t i o n s h i p , m y i m p r e s s i o n d u r i n g the s t u d y was t h a t p a t i e n t s are v e r y enthusiastic a b o u t this t e c h n o l o g y and w o u l d gladly pass up several office visits. Dr. T h o m a s F. O ' D o n n e l l (Boston, Mass.). I w o u l d like to make j u s t a s h o r t c o m m e n t , Dr. Bredenberg, at least at N e w England Medical Center w e ' r e n o t setting up telemedicine units in "distant" Maine b u t rather in less- r e m o t e places like Argentina, U n i t e d Arab Emirates, and Saudi Arabia. (laughter) I have one question for the authors a b o u t cellulitis and erythema. Obviously visualization is one o f the four com- ponents o f a physical examination, b u t telemedicine does n o t let you touch the w o u n d and sense the skin temperature or degree o f induration. You d o n ' t get this i m p o r t a n t aspect o f an examination o n a picture so that the examining physi- cian is unable to determine whether the skin is warm, which indicates cellulitis. D o yon want to c o m m e n t on that? D r . W i r t h l i n . V i e w i n g a d i g i t a l i m a g e in t e r m s o f d e t e c t i n g cellulitis and erythema will n o t be as g o o d as the standard physical examination. Also, we o b s e r v e d t h a t dig- ital images actually overrepresent r e d tones and t h a t fur- ther d e v e l o p m e n t in the t e c h n o l o g y is n e e d e d to improve the accuracy o f detecting e r y t h e m a and cellulitis. Dr. Jens J o r g e n s e n (South Portland, Me.). Maine is a big state, and there are some corners o f our state that are f u r t h e r f r o m P o r t l a n d t h a n N e w City is. A b o u t once a week I will see a patient who has driven 4 to 6 hours to come in for an appointment. Therefore if I get a call in the postoperative p e r i o d from the p a t i e n t or their physician that they are a little bit concerned a b o u t the appearance o f the w o u n d or the foot, it is oftentimes with a great deal o f reticence t h a t I say, " W h y d o n ' t you swing by the office and I ' l l take a look". So I think this technology has a great deal o f application in geographically disparate states such as Maine. I t h o u g h t it was a great presentation, and I w o u l d like to thank you for bringing this material to this forum. Dr. J o d A. B e r m a n (Springfield, Mass.). I think that the issue here i s n ' t so m u c h whether this t e c h n o l o g y can supplant the physical examination o f the p a r t o f the physi- cian, b u t rather its application in comparing the evaluation o f the visiting nurse with the evaluation o f the surgeon on the basis o f the images t h a t you obtain. Have y o u given any c o n s i d e r a t i o n t o c o m p a r i n g the accuracy o f t h e description t h a t the visiting nurse gives y o u with the eval- uation o f the physician on the basis o f y o u r images? Dr. W i r t h l i n . We plan to evaluate t h a t issue in the next phase o f the study.