E-Mail-Based Oral Medicine Consultation

A pilot study was undertaken to assess whether text-based electronic patient data transmission (e-mail) is a reliable source of information for the diagnostic decision-making process. The main objective was to determine if information contained within a transmitted text could be reliably used as basis for making general recommendations for diagnostic tests and follow-up or referral plans pertaining to a variety of oral mucosal pathologic conditions. The results suggest that face-to-face patient examination is more accurate in establishing a correct diagnosis for oral mucosal pathologies than transmitted descriptive patient data alone.

The explosive growth of the Internet and World Wide Web over the past few years has created great opportunities for rapid exchange of information online. With the Internet’s use having surpassed 60 million people,¹ its potential in facilitating communication within the health care industry cannot be ignored. An increasing number of physicians and institutions maintain Web sites and post e-mail addresses for a wide range of communications, including patient consultations.² In fact, over the past few years, e-mail communication has facilitated individual patient contacts with physicians, while telemedicine in general has made possible rapid communication between remote sites and major health care facilities. Most telemedicine programs in the United States support either interactive video-medical consultations or store-and-forward technology, with teleradiology being the most common application.³ Telemedicine has the potential to replace more-costly consultative services for hard-to-access areas such as rural communities and prisons,^4-7 as well as for home health care services.³

Despite the massive growth of telemedicine and its application to telecare, very little information is available in regard to its accuracy, cost-effectiveness, level of patient satisfaction, confidentiality issues, and licensing requirements. Since 1993, the Health Care Financing Administration has had studies in progress to assess the feasibility, acceptability, cost, quality, and access to services available through Medicare reimbursement of teleconsultations.³ Several other studies have focused on the quality of care issues related to telemedicine from both the medical and consumer standpoints.^8-10 There are no studies on the utility of telemedicine for dental consultations. Although online consultation may have application to many dental specialties, one logical application may be in the area of oral medicine. In the absence of published studies that can validate the use of teledentistry, information available in the medical field may be used until such studies become available in dentistry. Because of similarities between the diagnostic process and the management strategies of dermatology and oral medicine, published reports on telemedicine in dermatology may be used as models to assess the usefulness of an online oral medicine consultative service.^10-19

As the first step in developing an Internet-based oral medicine consultation service, a pilot study was undertaken at the University of California at Los Angeles Oral Medicine Department to assess whether text-based electronic patient data transmission (e-mail) is a reliable source of information for the diagnostic decision-making process. Although, e-mail-based communication may be useful for expert consultations in several areas of oral medicine -- such as management of medically compromised patients, orofacial pain, and temporomandibular joint diseases -- the authors focused this study on the diagnosis and management of oral mucosal diseases. Therefore, the main objective of the study was to determine if information contained within a transmitted text could be reliably used as basis for making general recommendations for diagnostic tests and follow-up or referral plans pertaining to a variety of oral mucosal pathologic conditions.

Study Objectives

To test the reliability and the accuracy (validity) of a diagnosis based on text-based information transmitted through e-mail.

Methods

Using the existing Oral Medicine Clinic patient database, 100 charts were randomly selected for the projects. The chart numbers were selected from a computer printout of the Oral Medicine Clinic patients list covering a five-year period, 1994-1999. These patients had been seen each time by a team of two faculty consisting of oral pathologists or oral medicine specialists attending the clinic. Two postgraduate residents from the orofacial pain program were asked to retrieve the charts and transfer all the relevant information from the charts to a standard patient data form previously designed for on-line consultations. The form consisted of a health history as well as the presenting signs and symptoms. To obtain consistent data on each patient, as well as the lesion in question, the form was designed to contain standard choices for each characteristic category (e.g., color, texture, comparison to the surrounding tissues, frequency, duration). Table 1 lists the contents of the patient data form. The sources of information used by the residents included patients’ registration forms, the health questionnaires, the faculty’s SOAP notes and the results of all appropriate diagnostic tests available in the charts (e.g., radiographs, blood workups, tissue biopsies, cytological smears). Each form was then coded by a unique identifier number and given to two oral medicine faculty judges (the authors) for review. The judges were blinded to the identity of patients and did not have access to any images from the patients. Using mainly the descriptive information contained in each form, each judge rendered two opinions about each patient: the most likely diagnosis and a recommendation consisting of a diagnostic test or a type of management and/or follow-up (mostly a broad category such as observation, excision, corticosteroids, or antifungal treatment).

Table 1 Patient Information Transfer Form

Patient Information Patient identifier (first initials only) Age Sex (M/F) Married (M/D/W) Descriptive History of the Problem Chief complaint? (in patient’s words): Describe patient’s symptoms: · Constant · Intermittent · When eating · When sleeping · Upon waking List all significant past and present medical conditions: List all major illnesses, hospitalizations, and allergies: List all current medication: List all medications taken during the past year: List relevant psychosocial history and habits: · Psychiatic conditions: · Social conditions (loss of job, homelessness, major personal problems): · Habits (tobacco, illicit substances, alcohol, or other): Has there been any treatment done for the conditions? · No · Yes (describe) If treated, has the problem stayed the same, gotten worse or better? · Same · Worse (Describe) What makes the problem worse? · Eating · Drinking · Brushing teeth or mouth wash · Touching area · Stress · Other: Other accompanying signs or symptoms? · Skin: · Eyes: · Nose: · Throat: · Other (describe):

Results of Patient Examination Indicate site with "X", mark on diagram, and describe Lips Labial mucosa Buccal mucosa Mucobuccal fold Gingiva (buccal, lingual, tooth or tooth site) Alveolar mucosa Tongue (dorsal, R or L lateral, R or L ventral, base), palate Tonsilar area Posterior pharyngeal wall Floor of the mouth R or L parotids salivary glands R or L submandibular salivary glands Lymph nodes (submandibular, anterior cervical, posterior cervical, buccal, other) Neck (midline, R o L lateral, posterior) Facial skin Hairline Other: Description of the lesion Color: Appearance: · Raised (papular) · Flat (macular) · Fluid filled (vesicular): · Papillary (verrucal): · Ulcerative: Size (in millimeter): Consistency: · Firm · Soft · Fluctuant · Not different than surrounding mucosa or skin Pain on palpation: Other information: Constitutional symptoms? Fever Malaise Lymphadenopahty Loss of appetite Recent weight loss None Image forwarded? Yes (format): No Radiographic information? Not relevant Relevant and as follows: Tooth Vitality Information? Not relevant Relevant as follows: Any Other Information?

All the forms were then returned to the residents, who independently scored them for agreement between the two judges as well as the accuracy of the opinions against the result of previous clinical examinations and confirmatory diagnostic tests (the gold standard).

To assess the level of agreement between each judge and the gold standard, as well as the level of concordance between the two judges, kappa analysis was used for both the "likely diagnosis" and the "recommendation" categories. Kappa analysis measures the degree of concordance between two observers while controlling for chance agreement. A kappa statistic of 0 implies levels of agreements that are completely attributable to chance alone. The closer the kappa value gets to 1.0, the higher the level of concordance between two measures unrelated to chance.

Results

A total of 78 charts from the original 100 had enough clinical information to be included in the study. Therefore, 78 lesions with intra- and extraoral presentations were documented and submitted to the judges for evaluation. Tables 2 and 3 (See printed version of the Journal for Tables 2 and 3.) list all the major categories of lesions found in the sample and the level of diagnostic agreement between each of the two judges and the gold standard (diagonal boxes from the upper left to the bottom right). For the ease of analysis, whenever a diagnosis was not relevant (i.e., normal anatomy) or was rare, the designation "Other" was used in coding the lesions.

The accuracy of diagnoses standard error (SE) for judge 1 and judge 2 was 58 percent 5.6 SE and 64 percent 5.4 SE respectively. The level of concordance between each judge and the gold standard was 0.60 (judge 1) and 0.53 (judge 2) kappa. The reliability (interjudge agreement) of the diagnoses made was shown to be 78 percent 4.7 SE with a concordance level of 0.76 kappa. The accuracy of the recommendations against the gold standard for judge 1 and judge 2 was 81 percent and 73 percent respectively. The level of agreement between the judges for a management plan (i.e., a biopsy or a similar laboratory diagnostic test, observation or pharmaceutical management) was 67 percent with a level of concordance of 0.43 kappa.

Discussion

The published information available on e-mail-based medical communications is limited. From the anecdotal reports in the literature, there seem to be an expansion of patient-physician e-mail contacts. Although there are no official reports of the number of physicians supporting a Web site, the impression is that the number is ever increasing.¹⁸ A physician Web site’s main application may be direct e-mail access for transmission of electronic patient history forms, appointment schedules, referrals, patients’ health reports (e.g., children’s camp forms), and, most importantly, patients’ concerns and questions.¹⁸ An increasing number of hospitals and health plans also support Web sites for introducing their personnel, services, and resources, including e-mail contacts. Internet access to major health centers may facilitate not only patient inquiries but also professional consultations among various providers across the country and worldwide. In one retrospective chart review of all e-mail consultations received by the Children’s Medical Center at the University of Virginia at Charlottesville, it was shown that of 1,001 consultations requested during a 33-month period, 81 percent were initiated by the children’s parents or guardian, 10 percent by physicians and 9 percent by other health professionals (nurses and pharmacists).²⁰ In the same study, it was shown that the consultation requests were received from 39 states and 37 other countries.

Telemedicine services are expanding and exploring new areas of patient services such as home health care. Projections are that home health care will change to a point where:

* An electronic stethoscope enables a physician or a home health nurse to listen to the heart and lungs of a patient with congestive heart failure;

* A pulse oximetry and respiratory flow data may be electronically transmitted to a physician for better management of a patient with chronic obstructive pulmonary disease;

* Visual observation of the insulin injection will improve the blood glucose monitoring of a poorly controlled diabetic; or

* Fetal heart monitoring of an obstetric patient becomes possible at home.³

There are many advantages to establishing telemedicine services. Telemedicine can offer a unique opportunity for improved access to specialty services by medically underserved areas. Starting this year, recognizing the growing use of telemedicine, the Health Care Financing Administration is reimbursing physicians for some telemedicine consultative services to Medicare patients. Through expanded reimbursement for services, individual practitioners or health centers that are located in rural communities can obtain rapid Internet access to specialty services in larger cities. Reports from several of these types of telemedicine programs, using mostly a video or store-and-forward technology, show rather high levels of diagnostic accuracy, patient satisfaction, and cost-effectiveness.^6,7,13 The diagnostic accuracy of teledermatology, using a captured image vis-à-vis a face-to-face examination, has been reported to range from 57 percent to 83 percent.^10-19 Burgiss and colleagues found that compared to the actual referral of a skin lesion for diagnosis and management by a dermatologist, the cost of obtaining a dermatology consultation via a telemedicine line and subsequent follow-up by the primary care provider was 37 percent lower.⁶ Recently published reports have also indicated that teledermatology consults significantly reduce the duration of a lesion prior to receiving an appropriate treatment.^6,21 One group showed that the average length of time for referral to a dermatologist was 13 weeks, while a teledermatology consultation was accomplished within 18 days.²¹ Clearly, shortening the duration of a dermatological condition has significant implications in early detection of premalignant or malignant lesions.

Despite all the advantages of telemedicine consultations (e.g., access to expertise), one must also note that there are still many legal and ethical issues to be considered. The legal ramifications of electronic transfer and storage of confidential patient information, the record-keeping requirements, the patient’s informed consent of electronic consultation, the licensing requirements, and the legal liabilities of the consulting specialists are all important considerations that have not been fully explored.²² In addition, the qualification requirements of the "consulting expert" must also be established.

In developing a telemedicine consultation service, the validity of the diagnoses made based on the electronic patient data transmission must be established. Depending on the complexity of the data transfer equipment used by the health care professional requesting the consultation, the information transmitted may contain descriptive patient data, digitized images or both. The pilot study described in this paper was designed as the first step in developing an oral medicine consultation service at UCLA Oral Medicine Department. The study’s objective was to test a patient data collection tool, designed for text-based consultation requests for oral mucosal pathologies submitted to the program’s Web site, but without any transmitted images of the patient. Although only one report had included the kappa calculation in its study (shown to be 0.3),¹⁷ the diagnostic agreement with the face-to-face examination reported by the dermatology literature (ranging from 57 percent to 83 percent, as described earlier)^10-19 is somewhat higher than the rate measured in this study (58 percent to 64 percent with the kappa equaling 0.5 to 0.6). It must be noted that the dermatologic studies published so far have all incorporated still or live images of the lesion in the diagnostic process. Therefore, the moderate diagnostic accuracy shown in this study, in all likelihood, is related to the lack of any visual information for the lesions. When adequate diagnostic information was available, the level of agreement between the two judges was high (78 percent with the kappa equaling 0.76). Although the level of agreement for the recommendations between the two judges and the gold standard was relatively high, the in-between judge concordance was low 67 percent (the kappa equaling 0.43). This may in part be related to differences in individual practitioner’s clinical experiences and management styles. For instance, one judge recommended a biopsy or a cytology procedure in 67 percent of cases, medications for 13 percent of lesions, observation for 12 percent of cases, and dental intervention (prosthodontic or restorative care) for 8 percent of lesions. The same categories for the other judge consisted of 55 percent biopsies or cytological examination, 23 percent medications, 18 percent observations, and 4 percent dental interventions.

The results of this pilot study suggest that face-to-face patient examination is more accurate in establishing a correct diagnosis for oral mucosal pathologies than transmitted descriptive patient data alone. Any general recommendations for a likely diagnosis or a type of management can be made at a moderate level of accuracy at best. Future studies focusing on the quality and reliability of intraoral image transfers, which should supplement text-based patient data, are necessary for establishing the accuracy of diagnoses made through teleoral medicine consultations. Based on the levels of diagnostic accuracy shown in this study, text-based information, without any transmitted images, should only be used for triaging of the consultations prior to the actual diagnostic decision-making. Until the time when adequate data transfer (including text and visual information) by the consulting practitioner is possible, the e-mail may be best used for exchanging ideas, disseminating the latest scientific information, and discussing the potential diagnoses only. It is clear that with appropriate data transfer equipment teleoral medicine may become a viable option for remote practitioners who require specialty consultations. The ultimate goal of a teleoral medicine consultation would be to enable the consulting provider (general dental practitioner, dental specialist, or medical provider) to make a sound patient management decision that may in some instances involve referral to an appropriate specialist.

Acknowledgement

The authors wish to thank Drs. Somsak Mitrirattanakul and Alan Stiles for their active participation in this study and Ms. Sylvia Swartz for her assistance with the patient chart sampling process.

Authors

Fariba S. Younai, DDS, is an adjunct associate professor of oral biology and medicine at the University of California at Los Angeles School of Dentistry.

Diana V. Messadi, BDS, DMedSc, is an adjunct associate professor of oral biology and medicine at the University of California at Los Angeles School of Dentistry.

References

1. Peters R, Sikorski R, Digital dialogue: sharing information and interests on the Internet. J Am Med Assoc 27:1258-60, 1997.

2. Neil RA, Mainous AG, et al, The utility of electronic mail as a medium for patient-physician communication. Arch Fam Med 3:268-71, 1994;.

3. Strode SW, Gustke S, Allen A, Technical and clinical progress in telemedicine. J Am Med Assoc 281:1066-8, 1999.

4. Murphy K, Telemedicine getting a test in efforts to cut costs of treating prisoners. New York Times June 8, 1998:D5.

5. Campbell NC, Ritchie LD, et al, Systematic review of cancer treatment programmes in remote rural areas. Br J Cancer 80:1275-80, 1999.

6. Burgiss SG, Julius CE, et al, Telemedicine for dermatology in rural patients. Telemed J 3(3):227-33, 1997.

7. Reid DS, Weaver LE, et al, Telemedicine in Nova Scotia: Report of a pilot study. Telemed J 4(3):249-58, 1998.

8. Houston MS, Myers JD, et al, Clinical consultations using store-and-forward telemedicine technology. Mayo Clin Proc 8:841-3, 1999.

9. Takeda H, Minato K, Takahasi T, High quality image oriented telemedicine with multimedia technology. Int J Med Inf 55:23-31, 1999.

10. Watts LA, Monk AF, Telemedicine: what happens in remote consultation. Int J Technol Assess Health Care 15:220-35, 1999.

11. Kvedar JC, Edwards RA, et all, The substitution of digital images for dermatologic physical examinations. Arch Dermatol 133:161-7, 1997.

12. Perednia DA, Gaines JA, Burtille TW, Comparison of the clinical informativeness of photographs and digital imaging media with multiple-choice receiver operating characteristics analysis. Arch Dermatol 131:292-7, 1995.

13. Oakly AMM, Astwood DR, et al, Diagnostic accuracy of teledermatology: result of a preliminary study in New Zealand. N Zeal Med J 110:51-3, 1997.

14. Gilmour E, Campbell SM, et al, Comparison of telecommunications and face-to-face consultations: preliminary results of a United Kingdom multicentre teledermatology study. Br J Dermatol 139:81-7, 1998.

15. Lowitt MH, Kessler II, et al, Teledermatology and in-person examinations. Arch Dermatol 134:471-6, 1998.

16. Whited JD, Mills BJ, et al, A pilot trial of digital imaging in skin cancer. J Telemed Telecare 4:108-12, 1998.

17. Philips CM, Burke WA, et al, Reliability of telemedicine in evaluating skin cancer. Telemed J 4:5-9, 1998.

18. Spielberg AR, On call and on line. J Am Med Assoc 280:1353-9, 1998.

19. Loane MA, Gore HE, et al, Preliminary results from Northern Ireland arms of the UK Multicentre Teledermatology Trial: effect of camera performance on diagnostic accuracy. J Telemed Telecare 3:73-5, 1997.

20. Borowitz SM, Wyatt JC, The origin, content, and workload of e-mail consultations. J AM Med Assoc 280:1321-4, 1998.

21. D’Souza M, Dhiren K, Ostler L, Dermatology opinions via Internet could reduce waiting times. Br J Med 318:737, 1999.

22. Spielberg AR, On line and on call. J Am Med Assoc 280:1353-9, 1998.



To request a printed copy of this article, please contact/ Fariba S. Younai, DDS, UCLA School of Dentistry, 10833 Le Conte Ave., Los Angeles, CA 90095-1668.