Disclosure
Teledermatology is best defined as the practice of dermatology using available communication and information technology. Given its visual nature, dermatology is well suited for telemedicine, and with advances in digital imaging technology and the availability of the Internet, the full potential of teledermatology (ie, improved patient access, more cost-effective medicine, continuing medical education and distance learning) seems more within reach. However, teledermatology cannot replace a traditional face-to-face dermatology evaluation in every instance. To date, studies on diagnostic agreement between teledermatology evaluations and in-person evaluations have reported a 59-92% correlation rate. This data is based predominantly on studies using live 2-way video teleconferencing (VTC) consultations. Studies evaluating the more practical solution, ie, store-and-forward (S&F) teledermatology, have been limited. However, the few studies evaluating diagnostic agreement with S&F teledermatology have been pilot trials using small sample sizes. Most recently, Whited et al reported 54% agreement among clinic-based examiners for their single most likely diagnosis and 92% agreement when ratings included differential diagnoses. Teledermatologists had statistically similar agreement in this study. According to a 2005 study by Mahendran et al, S&F teledermatology has limited diagnostic accuracy for skin lesions for skin cancer diagnosis and management. Du Moulin et al in 2003 reported that diagnostic categories with relatively high reliability of S&F teledermatology are eczema and follicular eruptions. Based on the literature, S&F teledermatology appears to be as clinically effective as VTC teledermatology. Furthermore, despite its limitations, S&F teledermatology has distinct advantages over VTC and will most likely be the predominant method used for teledermatology.
The 2 of methods teledermatology are S&F teledermatology and live VTC teledermatology. VTC uses a synchronous video and audio transmission to allow a live, interactive consultation between a patient and a remote physician (usually a specialist). S&F teledermatology is a method of sharing information in a manner independent of time (asynchronous) and place using digital images and clinical information received via the Internet for consultation with a remote dermatologist. Each mode has its advantages and disadvantages; however, most dermatologists tend to favor S&F teledermatology because it is less expensive and more practical. The major advantage of VTC teledermatology is that it most closely mimics the traditional face-to-face evaluation and allows a live interaction between the dermatologist and the patient. However, its distinct disadvantages are its higher cost, higher bandwidth, and, most importantly, requirement for coordination. On the other hand, S&F teledermatology is less expensive because it uses the widely available Internet and readily available commercial, off-the-shelf equipment (eg, digital camera). More importantly, S&F teledermatology does not require coordination between the physician and the patient, thereby saving valuable time.
The evolution of S&F teledermatology parallels the advances in digital imaging and the Internet, whose rapid growth has allowed fast and efficient transmission of digital information. Pioneering work by several dermatologists and recent advances in technology are responsible for the current state of teledermatology. Using 180 digitized slides of various dermatologic diseases at 3 different resolutions, Perednia et al reported that color slides and their digitized images (574 X 489 pixels at 24-bit color) were statistically similar. Using the multiple-choice receiver operating characteristic analysis, they found that this similarity was true even when stratified by difficulty of diagnosis. In 1997, Bittorf et al reported that images at 768 X 512 pixels at 24-bit color were perceived equivalent to images at higher resolutions. In an attempt to determine the effectiveness of imaging as a replacement for physical examination, Harrison used conventional photographic prints of skin tumors from 210 patients to make a diagnosis and compared the results with the criterion standard of histologic diagnosis. He reported a 71% diagnostic accuracy for the images studied by a dermatologist (histologic diagnosis after physical examination) compared with 49% accuracy for those studied by referring physicians (via prints). To test digital imaging in a real-world setting, Provost et al digitized and compressed (joint photographic E G [JPEG]-10-16:1 compression) conventional and dermatoscopic photo transparencies of 31 pigmented lesions (22 atypical melanocytic nevi, 9 biopsy-proven early malignant melanomas), sent them via the Internet, and then had dermatologists evaluate them via a color computer monitor. He determined an overall 87% rate of diagnostic agreement for clinical images and 92% agreement for dermatoscopic images. Although the dermatologists' performances in the diagnostic assessment of digitized images appear to not be affected by the level of compression (up to 40X), Sneiderman et al reported in 1994 that ratings of image quality for compressed images are significantly lower than for uncompressed images and for original photographic slides. With the reduction of image quality, Herrmann et al report that the reliability of the diagnosis is reduced. In 2005, Yagi et al and Brauchli et al report that new scanning technologies allow telepathological, entire-slide image analysis and conferencing in superior quality. However, standardization is lacking. |
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
What must be determined is how teledermatology compares with the criterion standard, whether this is the face-to-face dermatology evaluation or the histologic diagnosis. Most of the research to determine any discrepancy between teledermatology and other, standard methods has been derived from studies using interactive VTC consultations. Various authors have reported that the diagnosis made via teledermatology consultation is the same as that made via a face-to-face dermatology visit in 59-88% of cases. Although recognized that a dermatologist's diagnosis based on a face-to face evaluation cannot be 100% accurate, most studies addressing the reliability of teledermatology have compared interobserver diagnosis (teledermatology vs face-to-face consultation) to determine its effectiveness. Two previous studies, in part, used intraobserver diagnostic agreement and reported 71-82% agreement. However, the studies were performed using live VTC not S&F consultations. Pak et al reported the first diagnostic correlation study of an S&F teledermatology consultation system that used strictly an intraobserver comparison. This intraobserver study was an attempt to evaluate a teledermatology consultation system without taking into account the diagnostic variability among the dermatologists. Diagnostic agreement using S&F technology has been reported to be 61-91%. Variations in the results from these studies are due, in part, to sample size, the definition of agreement, interobserver variability, and the exclusion of certain types of skin conditions. Although they found no statistical difference in the disagreement rate between a face-to-face evaluation and a teledermatologist, Lesher et al reported that teledermatologists tended to give more broad differentials rather than a single diagnosis. Contrary to Kvedar et al, who reported that no specific disease category was more or less difficult to diagnose, Pak et al found that papulosquamous conditions, as a category, had the lowest rate of complete agreement (59%). This is consistent with the findings of Zelickson and Homan, who determined that eruptions were more difficult to diagnose than lesions. Loane et al also reported that eczematous conditions accounted for more than one third of all inappropriate management decisions. Although not completely clear, this lower diagnostic correlation and higher uncertainty may occur for several reasons, such as the inability to perform in-office tests (eg, potassium chloride staining) and to palpate the lesions. Perhaps more important is the potential sampling error, in which the contributing or referring physicians submit images that are not representative of the patient's true skin condition. Diagnostic confidence level In a study by Kvedar et al, the diagnostic certainty level (diagnostic confidence) correlated well with the type of diagnostic agreement. Not surprisingly, Pak et al found the diagnostic certainty level to be generally lower in the teledermatology group (confidence level = 7) when compared with the in-person evaluation group (confidence level = 9). The authors then hypothesized that the lack of confidence in the new computer technology, an inability to obtain additional history, an inability to perform in-office tests, and a lack of significant training in teledermatology may have accounted for the lower confidence level in the teledermatology group. A 2004 study by Whited et al showed a high level of satisfaction among all users of an S&F teledermatology consultation system. Most referring clinicians (92%) and dermatologist consultants (75%) reported overall satisfaction with the teledermatology consultation process. Teledermatology patients reached a point of initial intervention significantly sooner than did patients in usual care. Biopsy rates According to Shapiro et al, S&F teledermatology may provide an accurate and cost-effective method of determining whether patients with pigmented lesions and skin neoplasms presenting to primary care physicians should undergo biopsy. Given the overall lower diagnostic certainty levels of teledermatology consultations, higher biopsy rates would be expected in this group. According to the study by Pak et al, teledermatology evaluations recommended biopsies more often (10% more often vs in-person evaluations). However, most reports to date indicate no differences in the recommendations for biopsies between in-person evaluations and teledermatologist evaluations. This increase in the biopsy rate reported by Pak et al may be clinically significant because it may represent an increase in morbidity and, subsequently, increase the cost of dermatologic care. However, the authors suggested that this increase in the biopsy rate may be transient and will likely normalize as training improves and dermatologists become more familiar with the new consultation system. Finally, given the small number of biopsy specimens studied, concluding that teledermatology produces higher morbidity because of increased biopsy rates would be premature at this point. Limitations S&F teledermatology has certain inherent limitations. These include the quality of the image, the skill of the photographer, the limited history of the patient, the lack of interaction with the patient, the inability to palpate lesions, and the inability to perform simple laboratory tests such as potassium chloride staining. Some of these limitations can be minimized by improving technology and training (image quality), while others cannot be corrected. Despite these limitations, the data from the literature indicate that S&F teledermatology consultation systems seem to be clinically effective in the appropriate setting. In an evaluation of an S&F electronic referral system from 1999, White et al indicated that patients were more accurately triaged in at least 50% of cases and that 25% of patients did not require an outpatient dermatological appointment. Pak et al also found that a teledermatology system reduced unnecessary visits to the dermatologists by more than 50%. Outcome data Although the consensus on the diagnostic accuracy and clinical effectiveness of S&F teledermatology appears to be growing, outcome studies are needed to measure cost effectiveness and patient outcomes. A few small studies have evaluated the cost effectiveness of teledermatology. Wooton et al reported that VTC teledermatology was more expensive than conventional consultations. This was confirmed by Loane et al, who also found that S&F teledermatology, although less expensive, was not as clinically helpful (n = 204, randomized controlled trial). S&F teledermatology was not cost-saving when compared with usual care using observed costs and outcomes in a 2003 study by Whited et al; it was cost-effective for decreasing the time required for patients to reach a point of initial definitive care. To the authors' knowledge, a definitive outcome study has yet to be published. Satisfaction Patient satisfaction with multispecialty interactive teleconsultations was high in a study from the United Kingdom and among 96 patients who were seen via the Kentucky TeleCare telemedicine network. Pak et al also found high acceptance among patients, referring physicians, and consultants. Interestingly, patients seem to benefit most from this type of consultation system because it affords them easier access to a dermatologist. Based on a survey, Pak et al also found that their teledermatology consultation system provided education to referring physicians while allowing for a more appropriate use of dermatology resources. Experiences with S&F teledermatology in Switzerland (via a system that linked >80 dermatologists in private practice and universities) clearly indicated that it improved workflow and increased patients' confidence in their dermatologist, which, according to Kuhnis and Milesi, had translated into a high satisfaction and acceptance rate among patients. Lastly, in Huston and Burton's study of telemedicine, including psychiatry, dermatology, clinical nutrition, anesthesia, infectious diseases, rheumatology, internal medicine, neurology, and pediatric pulmonology, most patients were satisfied and only 16% stated they would have preferred to see the specialist in person.
The simplest S&F telemedicine systems use e-mail with images included as attachments, which can be generated and sent via widely available software such as Eudora or web browsers such as Internet Explorer or Netscape Communicator. An example of e-mail–based teledermatology is a proprietary system called DORIS in Norway. More advanced consultation systems have database retrieval and storage capability along with advanced features to facilitate the management of consultants. An example is the Walter Reed Army Medical Center's web-enabled S&F consultation system that is now in operation in more than 100 military bases in the United States and throughout the world. Because S&F teledermatology consultations are not yet reimbursable in most countries, little commercial effort has been expended to develop a robust system for a large organization. In fact, most efforts have been led by governmental or university-based organizations. Conforming standards for a teledermatology system are being developed by several organizations in various countries. Teledermatology in Switzerland In Internet-based telemedicine, data security is of uppermost importance because accessing unsecured data via the Internet is easy. Security measures built into the system also ease handling because no additional encoding of data is necessary. Without adequate security, medical information may easily fall into the wrong hands, and the patient may not consent to telemedical activities. The following hardware and configuration has been used in the Swiss teledermatology network, dermanet:
The communication platform used by dermanet is based on the Internet Protocol (TCP/IP). All services and transfers are password-, firewall-, and software-protected by Arpage's Security and Access System security software, which provides the following:
Authentication is achieved by keys of 1024-bit to 2048-bit (password or "passphrase" consisting of up to 256 characters). Encryption of data is based on 3DES with a 168-bit key. Data are also sent through a tunnel, further preventing unauthorized access and misuse of the information. This network, which has been used from 1995-2004, is under reconstruction. Teledermatological e-learning program SWISDOM/DOIT (Dermatology Online with Interactive Technology) is a free interactive e-learning platform for undergraduate and postgraduate students in dermatology. It is an online program to be used especially in countries with less developed infrastructure for teaching and medical care. It roughly consists of the following 4 parts:
The development and use of teledermatology has several obstacles, as follows:
Despite these obstacles to an S&F teledermatology system, the problems are not insurmountable and should not prevent the testing and development of a robust telemedicine consultation system.
Telemedicine will enable physicians to use modern telecommunications technology to gain access to specialists promptly and conveniently with satisfactory reliability and diagnostic accuracy. In the near future, telemedicine will become an integral part of the healthcare system. Much greater transmission bandwidth will be readily available, and advances in digital imaging technology (eg, CCD and compression technology) and in scanning histopathological slides for viewing, management, and analysis by virtual microscopy applications will result in image quality that is better than traditional film-based photography. The dermatological community must carefully evaluate the new technology and incorporate it into practice, where it has been demonstrated to be effective and efficient. To do this, a large-scale, multicentered randomized study evaluating patient outcomes (eg, quality of life, objective improvement) and cost effectiveness should be initiated. In addition, standards for telemedical referral must be devised for (1) content (eg, determining what specific information is required by the teledermatologist to make an accurate diagnosis) and (2) data structure (eg, HL-7–compatible, DICOM for images). Although many issues remain unresolved (eg, interstate medical licensing, liability, reimbursement, security, privacy), teledermatology will clearly enhance the delivery of dermatologic care to patients.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
){kind=small}
