NAVAL HEALTH RESEARCH CENTER A PROSPECTIVE EVALUATION OF ENT TELEMEDICINE IN REMOTE MILITARY POPU LA TIONS SEEKING SPECIAL TY CARE T. Melcer D. Hunsaker B. Crann L. Caola W. Deniston Report Document No. O1-4D Approved for public release; distribution unlimited. NAVAL HEALTH RESEARCH CENTER P 0 BOX 85122 SAN DIEGO, CA 92186-5122 BUREAU OF MEDICINE AND SURGERY (M2) 2300 E ST. NW WASHINGTON, DC 20372-5300 TELEMEDICINE JOURNAL AND e-HEALTH Volume 8, Number 3, 2002 © Mary Ann Liebert, Inc. A Prospective Evaluation of ENT Telemedicine in Remote Military Populations Seeking Specialty Care TED MELCER, Ph.D.,1 Capt. DARRELL HUNSAKER, M.D.,2 Commdr. BOBBI CRANN, R.N.,2 LISA CAOLA, M.A.,3 and Lieut. Commdr. WILLIAM DENISTON, Ph.D.1 ABSTRACT This study evaluated telemedicine use in remote military treatment facilities (MTFs) ashore over a 4-month period to help guide telemedicine applications for shipboard medical de- partments. A prospective study design was used to evaluate specialty care provided by an ear/nose/throat (ENT) physician via videoconferencing (VC) for patients at remote MTFs in TRICARE Region 9. The study provided a complete and continuous sample of ENT consul- tations during a planned 4-month period. Data sources included a telemedicine database and telephone interviews to assess attitudes of physician and nonphysician medical personnel. A total of 193 VC consultations (hereafter referred to as teleconsultations) were conducted fol- lowing referrals from primary providers. Patients were mostly young, male, active-duty per- sonnel. Forty-five percent of the 193 teleconsultations resulted in changed diagnosis by the ENT specialist relative to initial diagnosis by the referring provider. This rate of clinical im- pact was substantial, and it generalized across various ENT conditions, demographics, and MTFs. Medical personnel reported generally positive attitudes about telemedicine technolo- gies and the telemedicine process in TRICARE Region 9. Nonphysician providers reported slightly more favorable attitudes compared to physicians. These results suggest that ENT telemedicine has substantial clinical impact in the military populations treated at MTFs. A high rate of changed diagnoses (45%) was observed across age, gender, military status, ENT conditions, and treatment facilities. Medical personnel reported positive attitudes about us- ing the telemedicine system. These results support the use of telemedicine in shipboard med- ical departments. INTRODUCTION medicine can provide timely access to special- IT ized care for patients and medical personnel on HAS BECOME INCREASINGLY CLEAR that effec- ships at sea with limited medical resources.6 tive application of telemedicine technologies Telemedicine provides many benefits,7-9 which such as email, the Internet, and live videocon- extend to personnel and their dependents ferencing (VC) is essential for the U.S. armed treated at Military Treatment Facilities (MTFs) forces to optimize their health care mission."- in medically underserved regions ashore.10, 11 This is especially true for personnel deployed This study evolved from a naval research di- or stationed in remote environments.,' 6 Tele- rective to determine how telemedicine could 'Naval Health Research Center, Field Medical Technologies, San Diego, California. 2Naval Medical Center, San Diego, California. 3MTS Technologies, Inc., Arlington, Virginia. 301 302 MELCER ET AL. support shipboard medical departments.' medical technicians, were conducted to assess Larger ships, such as carriers and amphibious their attitudes and experiences with Region 9 ships, have extensive medical departments, in- telemedicine. Some evidence indicates that cluding physician and nonphysician personnel, telemedicine attitudes (e.g., satisfaction) and to perform the functions of a small hospital experience are positively correlated with (e.g., surgery). Some ships have telemedicine telemedicine use.21 Thus, satisfaction with technologies such as the Internet, VC, and telemedicine and experiences with telemedi- computed radiology installed.12 Unfortunately, cine and related technologies were measured the few evaluation studies of shipboard tele- among physician and nonphysician medical medicine processes that exist are limited to personnel. these large ships.13 No organized telemedicine capabilities currently exist on small ships, such as cruisers and destroyers, to assist medical de- METHODS partments headed by nonphysicians, such as independent duty corpsmen (IDCs).6 Evaluat- This was a prospective evaluation of ing telemedicine networks currently used at tedicin Tesecsltationsithat wee con- MTFs ashore provides an opportunity to study ducted by an ENT specialty physician. The data telemedicine use by military medical personnel were extracted from a telemedicine database and patient populations; results may general- called the referral management system (RMS) ize to help guide applications for shipboard and from structured interviews with physician medical departments. Some have pointed out and nonphysician medical personnel. The that the field of telemedicine in general needs study was conducted over a planned 4-month more systematic evaluation studies.14"5 period and included patient consultations and This study followed up on a retrospective medical personnel at remote MTFs in TRICARE evaluation of the first 2.5 years of telemedicine Region 9 and Lemoore Naval Air Station. network operation in TRICARE Region 9.11 In Region 9, remote MTFs, such as Fort Irwin and Port Hueneme, access specialty care by physi- Observations came from two sources: (1) A cians at Naval Medical Center, San Diego via database of telemedicine consultations and (2) live VC. The retrospective study showed that structured interviews with medical personnel. telemedicine use increased with time (since op- erational) at various MTFs and the duration of Referral management system database operation of individual telemedicine special- The telemedicine consultations (n = 193) oc- ties, e.g., ear/nose/throat (ENT). In addition, curred between November 1, 2000, and March the retrospective data suggested that approxi- 1 2001, and were extracted from the database. mately half of ENT consultations had a clinical These werea ll aessoi m th da ry impact, measured by change indiagnosis. These were all sessions in which primary The present study extended the retrospective providers referred patients for consultations TepsonfsetuRdeyg ionx9tenemedicinei netwoways: with an ENT specialist for initial evaluation. study of Region 9 telemedicine in two ways: (1) These consultations were provided for patients Analysis focused on determining a reliable rate treated at treatment facilities in TRICARE Re- of clinical impact, namely whether diagnosis changed, for ENT telemedicine over a planned gion 9 and at Lemoore Naval Air Station. These 4-month f o eriod. ENT conditions can facilities comprise the functional telemedicine 4mnhstudy peid N odtoscnnetwork centered at NMCSD."1 Individual pa- pose a concern for shipboard medical depart- ntw cntred on oMore obsevions for ments,16-18 and little evaluation exists of ENT tients contributed one or more observations for teleconsultations in MTFs.19,20 The generality one or more medical conditions. of these procedures for different ENT condi- tions and patient populations is unclear. (2) In- terviews with military medical personnel, e.g., Medical personnel from the Region 9 general medical officers (GMOs), IDCs, and telemedicine network completed telephone in- TELEMEDICINE AND MILITARY SPECIALTY CARE 303 terviews. These were physicians (n = 8) and Referral management system nonphysicians (n = 7), 11 males and 4 females. Telemedicine consultations were managed The physicians consisted of 2 GMOs, 3 primary and recorded via a secure Web-based applica- care physicians, and 3 specialty physicians tion installed for Region 9. When a primary (child psychiatry, ENT and neurology). The provider at a remote facility determined that a nonphysicians consisted of 1 IDC, 4 medical patient required specialty consultation, the pa- technicians, 1 nurse, and 1 physician's assis- tient was referred to a telemedicine coordina- tant. Of the 15, 12 were on active duty and 3 tor. If the specialty care was not available at the were civilians. This was a targeted sample patient's site, the telemedicine coordinator based on advice of the telemedicine coordina- (e.g., medical technician) scheduled a telecon- tor in TRICARE Region 9, who identified those sultation with the ENT physician. The patient who worked on these telemedicine consulta- at the remote site then was registered in the sys- tions. Over 90% of those targeted were con- tem along with information about background, tacted and completed the interview. A few mil- te am nwg i mato about baground, itay prsonelhadtrasferedto notersymptoms, and primary provider diagnosis. itary personnel had transferred to another The ENT specialist at San Diego then accessed assignment and were unavailable. There were the referral management system for scheduled no more than a few telemedicine personnel at consultations and patient case information. each clinic and at least 1 and as many as 4 med- During the consultation itself, the telemedi- ical personnel surveyed worked in each of 8 cine coordinator presented the patient to the MTFs. There were approximately 25 telemedi- consultant via VC and executed relevant cine personnel working in Region 9 telemedi- ENT examination procedures. Following the cine during the study period. An exact number telemedicine appointment, the consultant en- is difficult to determine due to turnover and tered comments on diagnosis and treatment in part time assignments. the patient's record. The choice of ENT specialty for this study TRICARE region 9 telemedicine network was based on the fact that such consultations were the most common in TRICARE Region TRICARE provides healthcare for active-duty 9.11 Initial evaluations were chosen because personnel, their qualified family members, these are the sessions where the primary Civilian Health and Medical Program of the UUninifiofromrmede dSSltee rrvviiccees -eliegliigbibllee rerteitrireee, s, aann d tthhee pthriosv si..d.tuedr ya,n sdt osrpee-cainadli-sfto rdwetaerrdm, inpere doipaegrnaotisvies. Ionr famTihleie ste alenmd esduircviinveo rns eotwf aolrl ku niinf osromuethde srenr vCicaelsi.- ptiossttoupderyat,i ve, oroe r foflloorww-uapr ds espsieoonsp weerree eoxr- cluded because these sessions usually did not fornia includes 8 MTFs [Naval Medical Center, function to determine diagnoses. The ENT San Diego, Port Hueneme, Fort Irwin, Edwards telemedicine examinations depended on real- Air Force Base (AFB), Vandenburg AFB, Twen- time videoconferencing, unlike other special- tynine Palms, Los Angeles AFB, Lemoore ties. For instance, dermatology makes exten- NAS], most with limitations on medical spe- sive use of the store-and-forward process such cialization. NMCSD functions as the "hub" of tthheel iztteeloeemmeeddiicciinn e netnwetwoorrkk.. TThhee rreomthoet e MMTTuFobss aasg neomsiasi.l3ing of digital pictures to determine di- are the "spokes" at various distances from anss The following variables were extracted from NMCSD (121-320 miles). Although the the database: Lemoore facility is not part of Region 9, it was included because it is part of the functional 9 Patient demographics (age, gender, and mil- telemedicine network centered at NMCSD. The itary status) spoke facilities received telemedicine consulta- * Reason for consultation (initial, preopera- tions from the ENT specialist at the hub in San tive, postoperative follow-up, other) Diego. Such a "hub-and-spoke" telemedicine ° Patient and primary provider site network model has been applied frequently in ° Consultant medical specialization (ENT, recent years in many settings.8'22 neurology, psychiatry, rheumatology) 304 MELCER ET AL. * Primary provider and consultant diagnosis * Usefulness of technologies (provider and consultant comments were * Perception of telemedicine (e.g., "Telemedi- coded into ICD-9 categories) cine allows me to give better health care to * Telemedicine modality (VC or store-and-for- my patients"). ward) Data analysis The ENT consultations included a diversity The data analysis addressed the following of problems. All cases were assigned to one of the following five categories: hearing (e.g., tin- questions: nitus), sleep-related (e.g., sleep apnea), tonsil- related, upper respiratory (e.g., sinusitis), and 0What percentage of teleconsultations had rineltaetgeudmguepnptearrnyr es(oer.gy. , (es.kgi.n, sckyy s t). stTiThheesse, ccaatte- - cdlainniscsablt esnitgenpiifaiycpaonicee , namely a difference nin gories were chosen in consultation with the iagnosis between the primary provider and lead ENT spelceaidaElNiTsstp.e iaTlst.hThuu,s , ii t wwaas poosssiibblle ttoo sseeee 0 tWheh iEchN,T i fs apneyci,a ldisetm?ographic or patient con- the relative frequency of major ENT conditions ditions were associated with changed diag- in this population, and to determine variation noses? in clinical impact of teleconsultations among e What were attitudes and experiences of these types of cases. medical personnel regarding telemedicine? Interviews with medical personnel Were these attitudes and experiences related to telemedicine use? Did attitudes vary by The medical personnel completed brief tele- level of medical training (i.e., physician or phone interviews, approximately 10 minutes nonphysician)? in length, to determine their medical back- grounds and experience with technology, and Chi-square tests (p < 0.05) were conducted to assess their attitudes toward telemedicine. to test for the significance of associations be- These interviews were conducted at the end tween change in diagnosis and ENT condition, of the study period according to a structured gender, and military status. The sample sizes questionnaire format. The questions and inter- for other variables were too small to permit this view forms were based on those used in pre- test. vious work at Naval Health Research Center.13 Descriptive statistics are presented for the in- The format was influenced by Dillman's23 terview data. Preliminary trends are described methodology for mail questionnaires. for the interview data because of the small sam- The following variables were recorded from ples of 8 physicians and 7 nonphysician med- the interview forms: ical personnel. Thus, the comparisons between physicians and nonphysicians should be inter- * Medical positions (e.g., IDC, nurse, physi- preted with caution. cian) and specialization (e.g., dermatology, psychiatry) • Years experience with telecommunications RESULTS technologies (e.g., VC, Internet, e-mail) * Medical purpose for telemedicine use (e.g., A total of 193 ENT teleconsultations were confirm diagnosis, patient education) conducted following referrals from primary * Attitudes were assessed using a 5-point Lik- providers. Forty-five percent of these consulta- ert rating scale (e.g., 1 = very satisfied, 2 = tions led to a change in diagnosis by an ENT somewhat satisfied, 3 = undecided, 4 = specialty physician relative to the initial diag- somewhat dissatisfied, 5 = very dissatis- nosis by the referring provider. This rate of fied). Five to 7 questions were used to assess change generalized across different ENT prob- each of four different attitudinal variables: lems, age, gender, military status, and facility * Satisfaction with technologies site. * Satisfaction with modalities (e.g., sound, im- Medical personnel reported generally posi- ages) tive attitudes toward telemedicine technologies TELEMEDICINE AND MILITARY SPECIALTY CARE 305 and the telemedicine process in TRICARE Re- 100- gion 9. Compared with physicians, nonphysi- cians reported slightly more favorable attitudes 90- and used telemedicine more often and in more 80- ways. 70- Demographics t; 60- The sample consisted of approximately 70% active-duty adult males between 18 and 44 50 years old (Table 1) and 30% females and mili- tary dependents. Over 75% of the consultations 40 came from the Lemoore or Port Hueneme fa- 30- cilities. An interruption in certification of the ENT specialist by Fort Irwin and Edwards AFB 20- limited ENT telemedicine at these sites. The de- 10 mographic variables will be considered for pos- sible associations with rate of diagnosis change. 0 Diagnosis Diagnosis Changed Unchanged Clinical impact Almost half (45%) of the consultations pro- FIG. 1. Clinical impact of VC consultations. duced a change in patient diagnosis by the ENT specialist relative to the initial diagnosis of the primary provider (Fig. 1). lated), this study explored whether the overall Because the ENT cases included diverse rate of clinical impact (Fig. 1) was general or medical problems (e.g., hearing- and sleep-re- limited to certain types of ENT problems, de- mographic (e.g., age and gender), or institu- tional variables (e.g., military or dependent sta- TABLE 1. DEMOGRAPHICS FOR PATIENTS REFERRED FOR ENT tus). Table 2 shows rates of diagnosis change CONSULTATIONS VIA VC CONSULTATIONS__V IA VCas a function of these variables. Total consults n = 193 %= 100 Table 2 shows that the rate of diagnosis change for the ENT consultations did not vary Gender Male 133 68.9 by ENT problem, gender, age, military status, Female 58 30.1 and or facility site. No significant deviations Unknown 2 1.0 occurred among the major ENT categories Age, years (i.e., hearing, sleep, tonsil, upper respiratory 1-10 19 9.8 (sp 11-17 5 2.6 [X2(3) = 1.43, p > 0.05]. However, none of the 5 18-24 50 25.9 integumentary cases led to a changed diagno- 25-44 92 47.7 sis; larger samples would be needed to deter- 45-65 19 9.8 65+ 7 3.6 mine the reliability of this latter trend. Simi- Unknown 1 .5 larly, no significant differences were seen Military status within gender [X2(1) = 0.97, p > 0.05] or mili- Active duty 132 68.4 tary status [X2(1) = 1.05, p > 0.05]. Dependent 59 30.6 Unknown 2 1.0 It may be noted that active-duty cases and MTF those 18-64 years old showed rates of diagno- Lemoore NAS 110 57.0 sis change similar to those found in the overall Port Hueneme 48 24.9 Twentynine Palms 17 8.8 sample. The various facility sites also appeared Fort Irwin 15 7.8 to have similar rates of diagnosis change, but Edwards AFB 3 1.6 compliance problems and small samples sizes ENT, ear/nose/throat; MTF, military treatment facil- from some facilities did not allow significance ity; VC, video teleconferencing. tests. The lack of substantial variation by these 306 MELCER ET AL. TABLE 2. CLINICAL IMPACT OF VC CONSULTATIONS BY SECONDARY VARIABLES Diagnosis changed No change Overall summary n = 87 % = 45 n = 106 % = 55 Type of ENT problem Hearing/vertigo 23 46 27 54 Sleep 12 38 20 62 Tonsil 13 46 15 54 Upper respiratory 39 50 39 50 Integumentary 0 0 5 100 Gender Male 63 47 70 53 Female 23 40 35 60 Age, years 1-10 6 32 13 68 11-17 2 40 3 60 18-24 22 44 28 56 25-44 43 47 49 53 45-64 9 47 10 53 65+ 5 71 2 29 Military status Active duty 62 47 70 53 Dependent 23 39 36 61 MTF Lemoore NAS 48 44 62 56 Port Hueneme 26 54 22 46 Twentynine Palms 5 29 12 71 Fort Irwin 6 40 9 60 Edwards AFB 2 67 1 33 Totals less than 193 cases reflect 1 or 2 missing cases for certain vari- ables (e.g. 192 for age and gender, 191 for millitary status). VC, Video teleconferencing. variables supports the generality of the overall Past telemedicine use rate of diagnosis change. Table 3 shows the medical reasons for which swith medical personnel the physician and nonphysician medical person- Interviews wnel said they had ever used telemedicine. The Structured interviews of 8 physicians and 7 primary reason given for past telemedicine use nonphysician medical personnel from the was to confirm diagnosis. All but 1 of the 15 med- TRICARE Region 9 telemedicine network sup- ical personnel had at some point used each of the plemented the data on patient outcomes. telemedicine technologies to confirm diagnosis. TABLE 3. PAST USE OF TELEMEDICINE TECHNOLOGIES BY MEDICAL PERSONNEL Physicians (n = 8) Nonphsicians (n = 7) Telemedicine Diagnosis Diagnosis technology confirmation (%) Education (%) confirmation (%) Education (%) Telephone 100 38 100 76 Fax 86 13 100 40 Email 93 38 100 81 Internet 100 75 100 85 VC 100 75 100 62 Overall average 95 48 100 69 VC, video teleconferencing. TELEMEDICINE AND MILITARY SPECIALTY CARE 307 TABLE 4. CURRENT ACCESS TO TELEMEDICINE TECHNOLOGIES BY MEDICAL PERSONNEL Physicians (n = 8) Nonphysicians (n = 7) Technology Current access (%) Years used (M) Current access (%) Years used (M) E-mail 75 6.3 100 3.0 RMS 75 1.5 100 1.7 Internet 75 3.6 100 3.6 VC 75 2 100 1.7 RMS, Referral ManagementSystem; VC, video teleconferencing. The physicians and nonphysicians differed more likely to use telemedicine for their own somewhat in past use of telemedicine for edu- education and for technical support. cational purposes (patient or provider). The nonphysicians were consistently more likely Telemedicine activity and satisfaction than physicians to use telemedicine for educa- Table 6 summarizes medical personnel re- tional reasons. At least 70% of both groups had ports of overall patient load and telemedicine used the Internet to obtain medical informa- activity during the study period (last 4 tion. atvt uigtesuypro ls months). Table 7 shows scores for satisfaction, sto telemedicine technologies usefulness, and overall perception of telemed- Current access ticine technologies (e.g., e-mail, VC) and satis- Table 4 shows that all nonphysicians and faction with different modalities (images, most physicians in the study had current ac- sound, written text). cess to various telemedicine technologies. All participants reported well over 1 year of expe- Level of telemedicine activity rience with each technology. Physicians and Nonphysicians reported more telemedicine nonphysicians reported similar experience activity than physicians, as shown in Table 6. with technology. This is seen both in the number of telemedicine cases and in the ratio of telemedicine cases to all patient care (telemedicine cases/telemedi- Table 5 shows the reasons physicians and non- cine plus non-telemedicine cases). This was pri- physicians used telemedicine during the study marily because a nonphysician at each clinic period (last 4 months). One of the physicians was was tasked to run all telemedicine sessions excluded from this analysis for lack of involve- including ENT, neurology, and psychiatry. In ment with telemedicine (last 4 months). contrast, physicians other than the ENT spe- More than three-fourths of all telemedicine cialist used telemedicine infrequently in Region activity was initiated to confirm diagnosis. This 9.11 Substantial variability occurred among was true for both groups. However, differences both physicians and nonphysicians on these between physicians and nonphysicians were measures, but the sample sizes were quite observed. Physicians were more likely than small (n = 7, n = 7). To reduce the effect of ex- nonphysicians to use telemedicine for patient treme fluctuations, median scores are pre- education. In contrast, the nonphysicians were sented here. TABLE 5. RECENT TELEMEDICINE USAGE BY MEDICAL PURPOSE Confirm Patient Provider Technical diagnosis (%) education (%) education (%) support (%) Unknown (%) Physician (n= 7) 79 19 1 1 0 Nonphysician (n = 7) 77 4 9 9 1 308 MELCER ET AL. TABLE 6. TELEMEDICINE AcTrrvTY REPORTED BY MEDICAL PERSONNEL Telemedicine activity Physicians Nonphysicians (last 4 months) (n = 7) (Mdn) (n = 7) (Mdn) Telemedicine patients seen (TMED) 60 85 Total patients seen (TMED + non-TMED) 900 200 Telemedicine ratio (TMED/TMED + non-TMED) 7% 43% Satisfaction with telemedicine larly evident in reactions to sound and video All personnel reported consistently favor- images. Both groups were undecided on satis- able attitudes toward telemedicine across all faction with faxed images. However, the physi- types of questions, shown in Table 7. They cians tended to be more satisfied with e-mailed rated their attitudes on a 5-point scale, with images than nonphysicians. more positive attitudes indicated by lower scores (see Methods). Nonphysicians reported DISCUSSION only slightly more favorable attitudes than did physicians. Although the size of this trend was small, it was seen for each of the four attitudi- These results support several conclusions. nal variables (technologies, modality, useful- First, the ENT teleconsultations had substantial ness, and perception). Each of the four attitu- impact on diagnosis. Nearly half of these tele- dina cvnaisaibtees dof fve o sven consultations led to a change in diagnosis by different questions. Nonphysicians reported the ENT specialist at San Diego relative to the more favorable attitudinal scores on the ma- initial diagnosis by the primary provider at the jority of questions, remote site. Importantly, this finding does not For example, when answering questions vary by type of ENT condition, patient demo- about their satisfaction with different tech- graphics, and medical facility site. Second, nologies, the nonphysicians indicated greater medical personnel in the TRICARE Region 9 satisfaction than physicians on five of the seven had considerable telemedicine experience and technologies. In particular, nonphysicians were reported generally positive attitudes toward more satisfied than physicians with the re the telecommunications technologies in gen- ral Management System and videoconferenc- eral and the telemedicine process. Third, non- ing. Both physicians and nonphysicians had physicians had slightly more positive attitudes ing.Bot phnsdic inso nphsicanshad about telemedicine than did the physicians. equally favorable ratings for store-and-forward ab te teredict hg ndid t icians. and Internet. This was an interesting finding, but it is not re- Responding to questions about satisfaction liable, given the small sample size. with specific modes of transmitted informa- tion, nonphysicians rated three of the five modes higher than did the physicians. The dif- It seems unlikely that the telemedicine pro- ferential between the two groups was particu- cess itself (e.g., VC) confounded the ENT ex- TABLE 7. TELEMEDICINE ATTITUDE SCORES REPORTED BY MEDICAL PERSONNEL Physicians Nonphysicians Telemedicine attitudes (n = 8) (M)Y (n = 7) (M) Satisfaction with telemedicine technologies 1.8 1.7 Satisfaction with modaltiy quality 2.1 1.9 Usefulness of telemedicine technologies 1.6 1.5 Perception of telemedicine/healthcare 2.0 1.7 Lower attitude scores indicate more positive attitudes. TELEMEDICINE AND MILITARY SPECIALTY CARE 309 aminations and caused the specialist to change others did not? Is it possible to decrease the the diagnosis. The essential portions of the ENT rate of referrals based on the educational im- examinations required viewing video images pact of the consultation process for primary from endoscopy cameras to examine interior of providers? Will increased exposure to specialty ear, nose, and throat. This information would consultations by the primary providers who be the same for "in-person" or telemedicine. generate the consults (e.g., GMO, IDC) lead to Previous studies have established high relia- improved diagnostic skill? bility between ENT telemedicine and "in-per- son" examinations.20'24 Medical personnel, in- Attitudes and telemedicine experience cluding the lead ENT specialist, rated video quality very highly. Finally, the patients were The interviews with medical personnel re- presented by trained medical personnel using vealed generally favorable attitudes toward standardized protocols. Therefore, it is reason- telemedicine technologies and the telemedicine able to use diagnoses via VC as the "gold stan- process. Other telemedicine studies have fre- dard," as is standard practice in Region 9, and quently reported similar positive findings. It that differences in diagnosis in the present was suggested this may be a measurement study reflect the patient's condition, not the problem, namely that patients and providers mode of consult. For this reason, no "in-per- were not asked critical questions.25 son" control was included in the present study. The present questionnaire instrument was Forty-five percent of ENT consultations led designed to include critical questions. The rat- to a change in diagnosis. This finding validates ings of medical personnel were less favorable the substantial rate of diagnosis change (49%) in response to some of these questions, such as reported in the initial study of TRICARE Re- "Is the quality of care rendered through the use gion 9 telemedicine.11 The data in the previous of telemedicine inferior to that provided in per- study were difficult to interpret because of son?" Other questions that drew less favorable missing observations, which could have biased responses were related to satisfaction with au- the observed rate in diagnosis change. The dio and video and image quality transmitted present study provided a complete and con- by fax. tinuous sample of ENT consultations during a Because the more critical questions did pro- planned 4-month period. Nonetheless, the duce less favorable ratings, the overall positive present study was limited to only diagnoses trend appears to reflect actual satisfaction with and not outcomes.4 Unfortunately, these data telemedicine rather than a measurement bias. were unavailable in the database. Hopefully, a Also, the trend for physicians and nonphysi- follow-up study will detail the clinical impact cians to show different patterns of response of these ENT teleconsultations. suggests that the survey instrument had some Several questions arise for further research. predictive validity. This study has addressed Will present telemedicine procedures be cost several internal validity issues. Although, these effective aboard ships? This question is espe- data were not statistically significant both reli- cially pertinent because the types of ENT prob- ability and validity of the survey instrument lems in this study (allergies, sleep-related) were plausible. Further research is indicated, could influence shipboard operational effi- however, to rule out bias. ciency. Upper respiratory conditions and aller- The interviews showed that medical person- gies may be a special concern for small ships, nel, particularly physicians, were concerned due to the restricted environments on these that telemedicine might compromise patient vessels.16 Some consultations (e.g., tonsillitis- care in some ways relative to in-person treat- related) could determine the need for surgery ment. In contrast, most of them agreed that and evacuation. Will the evaluation of consul- telemedicine would optimize overall care for tations similar to those in this study, conducted patients. by other ENT specialists, extend the generality ENT teleconsultations in TRICARE Region 9 of the present findings? Why did some consul- clearly profited from a functional infrastruc- tations lead to a change in diagnosis while ture, namely the telemedicine coordinator,