528

TELEMEDICINE AND e-HEALTHVolume 12, Number 5, 2006© Mary Ann Liebert, Inc.

Original Research

Teleconsultation for Cardiac Patients: A Comparison Between Nurses and Physicians:

The SHL Experience in Israel

ARIE ROTH, M.D.,1 ORI ROGOWSKI M.D.,2 YIGAL YANAY, B.A.,3MAYERA KEHATI, M.D.,3 NOMI MALOV, R.N.,3 and MICHAL GOLOVNER, M.D.3

ABSTRACT

The absence of randomized studies on sufficiently large patient cohorts precludes the draw-ing of any firm conclusions on the comparative performance between nurses and physiciansin transtelephonic triage and consultations and in diagnostic and management decision-mak-ing. We conducted such a comparative study at the SHL telemedicine facility. This facilityalso provides face-to-face medical management for its subscribers by means of mobile in-tensive care units (MICUs) staffed by physicians. The outcome of calls that came between7:00 AM and 11:00 PM throughout the study year and that were handled at random by speciallytrained physicians (n � 15) or nurses (n � 35) were analyzed. Of 48,707 subscribers who ful-filled the study entry criteria 25,106 used the service at least once, producing 88,103 calls (81,817handled by nurses and 6,286 by physicians). Teleconsultations were sufficient for most of thecases (80.13%). There were no significant differences between the performance of nurses andphysicians regarding demographics (age, gender) and medical diagnoses of the applicants.The nurses’ performance and decisions were comparable to those of physicians with respectto teleconsultations, medically justified dispatches of an MICU, repeated calls to the centerand mortality during the week after the index call, although the duration of the physicians’telephone consultations was longer. Delegation of equal authority to nurses and physiciansin triage and consultation in telecardiology results in equivalent and highly satisfactory med-ical care in a system in which subscribers receive service orchestrated from a single center oftelecommunications.

1Department of Cardiology, 2Department of Internal Medicine, Tel-Aviv Sourasky Medical Center, Sackler Facultyof Medicine, Tel-Aviv University, Tel-Aviv, Israel.

3SHL Medical Services, Tel-Aviv, Israel.

INTRODUCTION

ADVANCES IN TELECOMMUNICATION technolo-gies used in healthcare over the last

decade have greatly expanded the possibilitiesof its application to a variety of services. In ad-dition to remote monitoring of vital signs,

home health agencies now provide diverse sys-tems for palliative care (hospice), rehabilita-tion, case management, chronic disease man-agement, “virtual” house calls, postsurgicalfollow-up, and more. People are living longerthan ever before in history, are better educatedand more comfortable with the use of these

new technologies. Most individuals want to re-main independent and live in their own homesfor as long as possible, even though more andmore members of our aging society suffer fromchronic conditions. While critical shortages ofpublic resources force healthcare providers todo more with less, consumer expectations forquality care and enhanced clinical outcomescontinue to rise unabatedly. It is, therefore, notsurprising that experts have recommendedgreater involvement of home health services inthe care of patients with chronic disease.1–4

This recommendation is based on a number ofconsiderations, among which is the presump-tion that home healthcare nurses will reducehospitalization rates by educating patients, in-creasing patient compliance, detecting clinicaldeterioration and even decreasing costs tohealth providers.5–8

The concept of nurse practitioners is well de-veloped in the United States and has begun tospread to other Western countries as well. Inthe United Kingdom, for example, nurse prac-titioners are being increasingly used as the firstpoints of contact in primary care. Several au-thors have reported that physicians and nursesperform equally well in triaging patient visitsto their clinics,9–13 and that each group takenseparately is efficacious in performing trans-telephonic triage,14,15 and consultations16,17 aswell as in making diagnostic and managementdecisions.18,19 In-depth comparisons betweenthe performance of physicians and nurses,however, are few, and the ones that are avail-able involve small numbers of participants.Our search of the literature failed to elicit anyrandomized studies in sufficiently large patientcohorts to draw any firm conclusions.

SHL is a telemedicine system currentlyserving over 65,000 subscribers. As part ofSHL’s ongoing self-assessment, we wanted toconfirm that our policy of equal delegation ofauthority to nurses and physicians in triageand consultation was in the best interests ofour patients. To that end, we conducted thisstudy to evaluate the decisions taken bynurses compared to those taken by physiciansin the same setting of this telemedicine sys-tem that employs a relatively large profes-sional staff to serve a very large patient pop-ulation.

MATERIALS AND METHODS

The system

The medical facility SHL, which incorporatestelecardiological and face-to-face professionalmedical assistance to its subscribers through-out Israel and the Jordanian West Bank terri-tories, has been in operation since 1987 and hasbeen described in detail elsewhere.20–23 Briefly,SHL operates 24 hours per day, 365 days peryear through a monitoring center managed byspecially trained registered and/or intensivecare nurses who are authorized to dispatch mo-bile intensive care units (MICUs) staffed byphysicians and paramedics to subscribers liv-ing within a radius of 30 km of the cities of Tel-Aviv and Haifa. In the event that an SHL MICUis not available or when the call originates fromareas beyond SHL’s MICU catchment area, anMICU from MDA (the Israeli Red Shield) or aregular MDA ambulance (i.e., without a physi-cian in attendance) is dispatched.

Subscribers (predominantly cardiac patients)as well as healthy individuals (Table 1) learnabout the system through advertisement andarrive independently or are referred by theirphysicians or other treating medical staff(home for the elderly, etc.).

Each subscriber’s complete medical file,which also includes a full 12-lead electrocar-diogram, is stored in a central computer and

TELECONSULTATION FOR CARDIAC PATIENTS 529

TABLE 1. DISTRIBUTION OF THE LEADING DIAGNOSES OF

THE 48,707 SHL SUBSCRIBERS INCLUDED IN THE STUDY

DURING JANUARY–DECEMBER, 2003

Diagnosis (%)

Past myocardial infarction 24Anginal syndrome (any class) 23Dysrrhythmia (tachybrady arrhythmia 15

paroxysmal and/or chronic)Chronic conduction disturbances 15Poststatus any cardiac surgery 11Poststatus coronary angioplasty 10Congestive heart failure 8Postcardiopulmonary resuscitation 2Pacemaker/AICD 2Noncoronary heart disease 9Coronary risk factors only 28None of the abovea 27

aIncluding individuals who were ostensibly healthyand entirely symptom-free.

AICD, automatic implantable cardioverter defibrillator.

the records are continuously updated. All themembers carry a cardiobeeper by which theycan transmit a 3- (I, II, III) or 12-lead electrocar-diogram via a regular household or cellular tele-phone. Once a name or identification numberidentifies the caller, the appropriate medical fileis immediately displayed on a computer screenfrom which the nurse or physician retrieves allthe pertinent facts in addition to taking a real-time transtelephonic anamnesis. After interpret-ing the newly transmitted electrocardiographicdata, and based on the integration of writtenprotocols, computerized algorithms and her/hisclinical judgment, the nurse/physician may ei-ther dispatch an SHL MICU, an MDA MICU orambulance, or instruct the patient according tothe established medical or behavioral protocol.If it is the nurse who answers the call, he or shemay consult with the physician who is eitherpresent at the center or on-call. He or she is per-mitted to recommend over-the-counter medica-tions or those used regularly by the patient, butnot new drugs.

The staff is capable of handling calls in He-brew, Arabic (Israel’s second official language),Russian, and English.

Staff training

In order to be accepted as a staff member atSHL, both physicians and nurses must take awritten examination to evaluate their expertisein straightforward common cardiological is-sues, in the interpretation of case studies, in theidentification of life-threatening situations, andin the fields of electrocardiography and phar-macology (the latter with special emphasis oncardiovascular drugs). The candidates whomeet the stringent requirements are inter-viewed by two experienced senior physiciansin order to assess their suitability (in terms oftheir character) for working in this special set-ting. The ones who pass these tests are inter-viewed by the chief cardiologist, and specificgaps in their knowledge as revealed by earliertests are addressed. The candidates who arefound to need it undergo SHL’s 160-hourcourse on executing and interpreting electro-cardiograms. The final stage of training in-volves 200 hours of hands-on experience in themonitor center under close supervision of a

personal instructor (an experienced nurse) and60 hours with the MICU team in the field. It isduring this period that the trainee becomeshighly familiar with all of the SHL protocolsand standing orders.

Continuing education and quality assurance

A staff meeting takes place every month forphysicians and every 3 months for nurses withthe purpose of analyzing and addressing newissues and orders, case studies and updates. Allcalls to and from the monitor center arerecorded, and at least 2–3 times per year threecalls are chosen at random for each staff mem-ber for analysis by the medical management (3physicians [including 1 cardiologist], 1 headnurse and 1 operational manager) who listento the calls and professionally assess the waythey were handled. The staff member is theninformed about the quality of his/her judg-ment and actions, and these remarks are addedto his/her personal file for comparison with thenext analysis.

Study period and patient population

After administrative and other miscella-neous calls had been screened out, all medicalcalls to the monitor center between January toDecember, 2003 were reviewed for this study.Only calls that came in between 7:00 AM and11:00 PM were included because those were thehours when physicians who triaged patientswere present at the monitor center.

DESIGN

Neither the nurses nor the physicians wereaware that their performances were being compared. The outcomes of all their telecon-sultations were assessed with respect to thesubscribers’ medical history that had been retrieved from their medical files. The endpoints for comparison and characterizationwere as follows.

Teleconsultations

These included all calls that were concludedon the telephone without any additional mea-

ROTH ET AL.530

sures, all calls which resulted in an additionalteleconsultation(s), and all calls that resulted inthe dispatch of an SHL MICU.

Justified dispatches of a MICU

All dispatches of a SHL MICU that resultedin the subscriber’s being transported to theemergency department of a hospital or in thenecessity for emergent or any medical treat-ment on the scene were considered as being“justified.” Of these, the ones in which the pa-tient needed to be taken to the hospital wereclassified as first order (Type A) justified calls,while the ones in which the dispatch concludedwith treatment of any kind at the scene (in-cluding as little as the administration of an oraltranquilizer) were considered as being secondorder (Type B) justified calls.

Unjustified dispatches of an MICU

All dispatches of an SHL MICU in which theexamining physician left the patient at homewithout any treatment or suggestion apartfrom the recommendation to call the centeragain in case of the reappearance or aggrava-tion of symptoms were considered as havingbeen unjustified.

Follow-up

Data on additional application(s) to the SHLmonitor center and on mortality during theweek after the index call to the center as wellas the result of dispatches handled by SHL’sMICUs were retrieved. Because information onthe outcome of dispatches by other than SHLstaff members could not be completely docu-mented, only those dispatches treated by theSHL team were included in the final analysis.Mortality data could, however, be obtainedfrom relatives who called SHL to cancel thesubscription of deceased members.

Statistical analysis

Data were downloaded from the database ofsubscriber calls to the SHL center and trans-fered into the Statistical Package for the SocialSciences (SPSS, Chicago, IL) for analysis. Re-sults are given as mean � standard deviation(SD) for continuous variables, such as staff

members’ working experience, and as numbersplus percentages for all dichotomous variables,such as gender, justified dispatches, etc. The �2

analysis was used to compare the differentvariables between nurses and physicians. A pvalue �0.05 was considered statistically signif-icant in the analyses of the results.

RESULTS

General

During the study period, 48,707 subscribers ful-filled study entry criteria and their main cardiacdiagnoses are displayed in Table 1. There were26,799 (55%) males, and the age distribution ofthe cohort was 47% who were 61–80 years of age;2%, 40 years or less; 25%, 41–60 years; and 26%,81 years or older. During the study period 25,106subscribers (52%) called the center at least onceand 1 patient called 815 times. The breakdown ofthe frequency of calls is shown in Table 2.

The SHL medical staff that participated in thestudy consisted of 35 registered nurses withprofessional experience of 84 � 60 months(range, 6–199 months) and 13 physicians withprofessional experience of 56 � 38 months(range, 18–139 months).

Teleconsultations and triage

A total of 154,796 telephone calls consistingof teleconsultations or the dispatching of an

TELECONSULTATION FOR CARDIAC PATIENTS 531

TABLE 2. NUMBER OF CALLS PER SUBSCRIBER

DURING THE ONE-YEAR STUDY PERIOD

No. of calls/year Subscribers (n � 48,707)

0 23,601 (48%)1 9,8622 5,2523 3,0464 1,9045 1,2746–10 2,43811–20 93221–30 20331–40 6541–50 6151–75 4776–100 20101–200 19201–600 8815 1

MICU or ambulance were recorded through-out 24 hours per day during 2003, includingthose of subscribers without known or docu-mented cardiac disease. Of these, 94,169 callscame in between 7:00 AM and 11:00 PM and werehandled randomly by either nurses or physi-cians (whoever was available and picked upthe telephone).

All calls flagged with a standing order forphysician consultation only or for mandatorydispatch of an MICU (e.g., direct admission re-quested by the subscriber’s private physician)were excluded. This left a total of 88,103 callsof which 81,817 were handled by nurses (aver-age, 2,338 each) and 6,286 by physicians (aver-age, 484 each).

Telephone consultations lasted 6.88 � 6.98minutes for nurses and 8.97 � 6.88 minutes forphysicians (p � 0.001). Those calls that endedwith an MICU dispatch tended to be a littlelonger for both nurses and physicians.

There were no significant differences be-tween the calls handled by nurses or physiciansin terms of the patient’s age, gender, distribu-tion of medical diagnoses, or the hour at whichthe call was received.

Dispatches

A dispatch of an MICU (SHL’s or other) or aregular ambulance resulted from 19.87%(17,502/88,103) of the calls, while the remain-ing 70,601 (80.13%) calls ended in teleconsulta-tions alone. Overall, nurses tended to dispatchMICUs slightly more than physicians (in 19%of the calls compared with 18%, respectively,p � 0.073). An SHL MICU was dispatched inthe majority of cases (approximately 70%,12,150 calls) of which 69.3% (8,423/12,150)were considered justified dispatches. Nursesdispatched more justified cases (67%) thanphysicians 64% (p � 0.049). A comparison be-tween nurses and physicians for type of dis-patch revealed that 50% of all the nurses’ dis-patches and 45% of all the physicians’dispatches were “Type A” (p � 0.001) while20% of all the physicians’ dispatches and 17%of all the nurses’ dispatches were “Type B” (p �0.01). Transportation to the hospital followedthe MICU team’s on-the-scene treatment for41% of the nurses’ dispatches compared with

only 35% of the physicians’ dispatches (p �0.001).

No correlation was found between the extentof the nurses and physician’s professional ex-perience and their tendency to dispatch anMICU.

Follow-up

Of the total number of calls, 29% (25,865/88,103) called again (by request or from their owninitiative for reassurance) within 24 hours fromthe index call. Of these, 29% (23,817/81,817) werecalls that had originally been handled by nursescompared with 33% (2,048/6,286) handled origi-nally by physicians (p � 0.001).

Of the 70,601 calls that ended with teleconsul-tations, a dispatch was required during the fol-lowing week in 9.08% of those handled by nursesand in 9.88% of those handled by physicians,with a subsequent mortality of 0.03% and 0%, re-spectively. None of these differences reached alevel of statistical significance (p � 0.25).

DISCUSSION

Telephone consultation is becoming an in-creasingly accepted approach to patient careand was shown to improve public access tomedical information and advice. Despite therapid advances and dissemination of informa-tion technologies in our daily life, however, theuse of telemedicine remains restricted ratherthan being widely deployed. Moreover, evenrecent systematic reviews of telemedicine,some addressing specific outcomes (e.g., im-pact on clinical decision making and clinicaloutcomes,17–19 patient satisfaction24,25 and costeffectiveness19,26), have provided disappoint-ingly little information about when telemedi-cine applications are practical and when theyare not. One of the concerns may involve theability of nurses to use telemedical applicationsfor appropriately triaging patients.

While the ability of nurses to triage patientsin stand-alone clinics6 or independently butstill under the roof of colleague or partnerphysicians13 is well documented, we believethe present study to be the first to show thatthe capabilities of well-trained nurses to triage

ROTH ET AL.532

and consult through the telephone are compa-rable to those of physicians. This investigationwas done on a large population of patients whowere handled at random by physicians andnurses who were unaware that their decisionswere being studied and compared. The SHLsystem has the advantage that applicants arenot anonymous and that each subscriber has afully updated medical record, thus providingessential information to the “triager,” com-pared to one who has to make decisions for acompletely anonymous patient.

We did not attempt to compare the clinicaleffectiveness of telemedicine with standardhealthcare methods nor the diagnostic accu-racy of the system versus conventional diag-nostic methods. Nevertheless, our results are inaccordance with those of others and providemore evidence for the utility of teleconsultationin cardiology.8,15 Noteworthy, the attitudes ofpatients towards the use of telemedicine havebeen measured by many authors, with most re-porting high levels of satisfaction and a will-ingness to use telemedicine in the future.24,25,27

We, as others found that a large proportion ofcalls (about 80%) can be handled with advicealone.6 In contrast to studies that comparedphysicians and nurse’s face-to-face consultationsand reported on longer consultation time fornurses,10,18,27 our experience was the opposite:the physicians’ consultations lasted significantlylonger than those of the nurses (8.97 � 6.88 min-utes versus 6.88 � 6.98, respectively, p � 0.001).Although it is tempting to attribute the longerconsultation time for physicians to an element ofuncertainty, we suspect that the reason lies intheir being less experienced with the computersoftware compared to the nurses.

Ten years ago, an anonymous editorial in TheLancet stated, “Although much is claimed, theeconomic benefits of telemedicine have yet to beproved.”28 It was beyond the scope of this paperto address economic issues, but our demonstra-tion of nurses’ triage capabilities shows promisein the system for reducing costs without jeopar-dizing patients’ health, especially in those areasthat lack sufficient numbers of physicians orwhere a nurse’s salary may be lower than a physi-cian’s. It should also be taken into account thattelecardiological services may be easier to applythan for other fields because it relies heavily on

analyzing information from a patient’s back-ground and history in assessing the patient’s condition, in addition to a reliable electrocardio-graphic transmission and a reliable electrocar-diographic transmission that does not need ul-trasophisticated image capture and resolutiondevices as in other fields of telemedicine (i.e., tel-eradiology, telepathology, teledermatology, etc.).

Finally, an important factor for the success ofany telecardiological program is the training ofthe staff. Nurse practitioners are increasinglyused as the point of first contact in primarycare, and observational studies have suggestedthat patients react positively to the use of nursepractitioners.27 As a result, the number oftrained nurse practitioners is increasing as ded-icated training programs become more accessi-ble and quality assurance is established to mon-itor their effectiveness.13–15,29

Our findings have demonstrated that equaldelegation of authority to nurses and physi-cians in triage and consultation in telecardiol-ogy results in equivalent and highly satisfac-tory medical care to the subscribers of thesystem. The fact that the services are orches-trated from a single center of telecommunica-tions means that the system can quickly reachboth surrounding and remote rural areas aslong as a telephone line is available.

ACKNOWLEDGMENT

Esther Eshkol is thanked for her superb edi-torial assistance.

DISCLAIMER

Dr. Roth is a medical consultant of SHL Med-ical Services.

REFERENCES

1. Stewart S, Marley JE, Horowitz JD. Effects of a mul-tidisciplinary, home-based intervention on plannedreadmissions and survival among patients withchronic congestive heart failure: A randomised con-trolled trial. Lancet 1999;354:1077–1083.

2. Rich M, Beckham V, Wittenberg C, Leven C, FreedlandK, Carney R. A multidisciplinary intervention to preventthe readmission of elderly patients with congestive heartfailure. N Engl J Med 1995;333:1190–1198.

TELECONSULTATION FOR CARDIAC PATIENTS 533

3. de Lusignan S, Wells S, Johnson P, Meredith K,Leatham E. Compliance and effectiveness of 1 year’shome telemonitoring. The report of a pilot study ofpatients with chronic heart failure. Eur J Heart Fail2001;3:723–730.

4. Harrison J, Marshall J. Psychiatric home treatment.Home treatment works. Br Med J 2000;321:177–178.

5. Malasanos TH, Burlingame JB, Youngblade L, PatelBD, Muir AB. Improved access to subspecialist dia-betes care by telemedicine: Cost savings and caremeasures in the first two years of the FITE diabetesproject. J Telemed Telecare 2005;11(Suppl 1):74–76.

6. Lattimer V, George S, Thompson F, Thomas E, MulleeM, Turnbull J, Smith H, Moore M, Bond H, GlaspurA. Safety and effectiveness of nurse telephone con-sultation in out of hours primary care: Randomisedcontrolled trial. BMJ 1998;317:1054–1059.

7. Lattimer V, Sassi F, George S, Moore M, Turnbull J,Mullee M, Smith H. Cost analysis of nurse telephoneconsultation in out of hours primary care: Evidencefrom a randomised controlled trial. BMJ 2000;320:1053–1057.

8. Roth A, Malov N, Carthy Z, Golovner M, Naveh R, Al-roy I, Kaplinsky E, Laniado S. Potential reduction ofcosts and hospital emergency department visits result-ing from pre-hospital transtelephonic triage. The “Sha-hal” experience in Israel. Clin Cardiol 2000;23;271–276.

9. Smith LE, Fabbri SA, Pai R, Ferry D, Heywood JT.Symptomatic improvement and reduced hospitaliza-tion for patients attending a cardiomyopathy clinic.Clin Cardiol 1997;20:949–954.

10. Venning P, Durie A, Roland M, Roberts C, Leese B.Randomised controlled trial comparing cost effec-tiveness of general practitioners and nurse practi-tioners in primary care. BMJ 2000;320:1048–1053.

11. Laurant M, Reeves D, Hermens R, Braspenning J, GrolR, Sibbald B. Substitution of doctors by nurses in pri-mary care. Cochrane Database Syst Rev 2005 Apr18;(2):CD001271.

12. Kinnersley P, Anderson E, Parry K, Clement J, Ar-chard L, Turton P, Stainthorpe A, Fraser A, Butler CC,Rogers C. Randomised controlled trial of nurse prac-titioner versus general practitioner care for patientsrequesting “same day” consultations in primary care.BMJ 2000;320:1043–1048.

13. Mundinger MO, Kane RL, Lenz ER, Tolten AM, TsaiWY, Cleary PD, Friedewold WD, Siu AL, ShelanskiML. Primary care outcomes in patients treated bynurse practitioners or physicians: A randomized trial.JAMA 2000;283:59–68.

14. DeVore NE. Telephone triage. A challenge for prac-ticing midwives. Nurse Midwifery 1999;44:471–479.

15. Chiantera A, Scalvini S, Pulignano G, Pugliese M, De-Liol, Mazza A, Fera MS, Bussolotti L, Bartolini S,Guerrieri L, Caroselli A, Giovannini E. Role of tele-cardiology in the assessment of angina in patientswith recent acute coronary syndrome. J Telemed Tele-care 2005;11(Suppl 1):93–94.

16. Marklund B, Bengtsson C. Medical advice by tele-phone at Swedish health centres: Who calls and whatare the problems? Fam Pract 1989;6:42–46.

17. Dale J, Williams S, Foster T, Higgins J, Snooks H,Crouch R, Hartley-Sharpe C, Glucksman E, George S.Safety of telephone consultation for “non-serious”emergency ambulance service patients. Qual SafHealth Care 2004;13:363–373.

18. Hersh W, Helfand M, Wallace J, Kraemer D, PattersonP, Shapiro S, Greenlick M. A systematic review of the ef-ficacy of telemedicine for making diagnostic and man-agement decisions. J Telemed Telecare 2002;8:197–209.

19. Roine R, Ohinmaa A, Hailey D. Assessing telemedi-cine: A systematic review of the literature. CMAJ2001;165:765–771.

20. Roth A, Herling M, Vishlitzki V, Aitkin I. The impactof SHAHAL’ (a new cardiac emergency service) onsubscribers’ appeal for medical assistance: character-istics and distribution of calls. Eur Heart J 1995;16:129–133.

21. Roth A, Golovner M, Malov N, Carthy Z, Naveh R,Alroy I, Kapinsky E, Laniado S. The ‘Telepress’ sys-tem: self-measurement and monitoring of blood pres-sure. The ‘Shahal’ experience in Israel. Am J Cardiol1999;83:610–612.

22. Roth A, Malov N, Golovner M, Slesarenko Y, Naveh R,Kaplinsky E, Laniado S. Assessment of a ck-m/myo-globin kit in the pre-hospital setting in patients pre-senting with acute non-traumatic chest pain. The ‘Sha-hal’ experience. Crit Care Med 1999;27:1085–1089.

23. Viskin S, Golovner M, Malov N, Fish R, Alroy I, VilaY, Laniado S, Kaplinsky E, Roth A. Circadian varia-tion of symptomatic paroxysmal atrial fibrillation.Data from almost 10,000 episodes. Eur Heart J1999;20:1429–1434.

24. Williams TL, May CR, Esmail A. Limitations of pa-tient satisfaction studies in telehealthcare: a system-atic review of the literature. Telemed J E Health2002;7:293–316.

25. Mair F, Whitten P. Systematic review of studies of pa-tient satisfaction with telemedicine. Br Med J 2000;320:1517–1520.

26. Whitten PS, Mair FS, Haycox A, May CR, WilliamsTL, Hellmich S. Systematic review of cost effective-ness studies of telemedicine interventions. Br Med J2002;324:1434–1437.

27. Taylor P. Evaluating telemedicine systems and ser-vices. J Telemed Telecare 2005;11:167–177.

28. Telemedicine: fad or future [editorial]? Lancet 1995;345:73–74.

29. Keatinge D, Rawlings K. Outcomes of a nurse-led tele-phone triage service in Australia. Int J Nurs Pract2005;11:5–12.

Address reprint requests to:Arie Roth, M.D.

Department of CardiologyTel-Aviv Sourasky Medical Center

6 Weizman StreetTel-Aviv, 64239

Israel

E-mail: arieroth@tasmc.health.gov.il

ROTH ET AL.534