Electronic Patient-PhysicianCommunication: Problems and Promise

The Harvard community has made thisarticle openly available. Please share howthis access benefits you. Your story matters

Citation Mandl, Kenneth D., Isaac S. Kohane, and Allan M. Brandt. 1998.Electronic patient-physician communication: Problems and promise.Annals of Internal Medicine 129(6): 495-500.

Published Version http://www.annals.org/

Citable link http://nrs.harvard.edu/urn-3:HUL.InstRepos:3382980

Terms of Use This article was downloaded from Harvard University’s DASHrepository, and is made available under the terms and conditionsapplicable to Other Posted Material, as set forth at http://nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of-use#LAA

MEDICINE AND PUBLIC ISSUES

Electronic Patient-Physician Communication:Problems and PromiseKenneth D. Mandl, MD, MPH; Tsaac S. Kohane, MD, PhD: and Allan M. Brandt, PhD

A critical mass of Internet users will soon enable widediffusion of electronic communication within medicalpractice. E-mail between physicians and patients offersimportant opportunities for better communication. Link-ing patients and physicians through e-mail may increasethe involvement of patients in supervising and document-ing their own health care, processes that may activatepatients and contribute to improved health. These newlinkages may have profound implications for the patient-physician relationship. Although the federal governmentproposes regulation of telemedicine technologies andmedical software, communications technologies are evolv-ing under less scrutiny. Unless these technologies are im-plemented with substantial forethought, they may disturbdelicate balances in the patient-physician relationship,widen social disparities in health outcomes, and createbarriers to access to health care.

This paper seeks to identify the promise and pitfalls ofelectronic patient-pbysician communication before suchtechnology becomes widely distributed. A researchagenda is proposed that would provide data that areuseful for careful shaping of the communications infra-structure. The paper addresses the need to 1) define ap-propriate use of the various modes of patient-physiciancommunication, 2) ensure the security and confidentialityof patient information, 3) create user interfaces that guidepatients in effective use of the technology, 4) proactivelyassess medicolegal liability, and 5) ensure access to thetechnology by a multicultural, multilingual populationwith varying degrees of literacy.

Arm Intern Med. 1998:l29:495o00.

From Children's Hospital and Harvard Medical School, Boston,and Massachusetts Institute of Technology. Cambridge, Massa-chusetts. For current author addresses, see end of text.

The introdnction of the telephone into medicalpractice in the last decades of the 19th century

was greeted with both celebration and trepidation.Invented in 1876 by Alexander Graham Bell, thetelephone was commercially introduced in the late1870s. It was not until World War I, however, thatthe telephone became a common utility (1). Al-though some physicians heralded the advantages ofefficiency and accessibility that the telephone of-fered, others expressed concerns about being over-whelmed by patients seeking over-the-telephonecare, the safety of telephone diagnosis, and prob-lems of privacy. By the mid-1920s, the telephonewas fully integrated into physician practice, as it was

in broader society. The telephone had become amandatory medical technology, as central to prac-tice as the stethoscope and sphygmomanometer (2).

We are again on the threshold of a dramaticexpansion in communications technology that mayhave profound effects on the patient-physician rela-tionship and the practice of medicine. We are ap-proaching a critical mass of Internet users that willlead to a wide diffusion of electronic communica-tions within medical practice (3, 4). The AmericanMedical Informatics Association recently publishedrecommendations to guide computer-based commu-nications between clinicians and patients (5). Littleattention, however, is being paid to the implicationsof direct electronic linkages between physicians andpatients. Unless implemented with substantial fore-thought, these linkages may disturb delicate bal-ances in the patient-physician relationship, widensocial disparities in health outcomes, and createbarriers to access. We attempt to identify some ofthe promise and pitfalls of electronic patient-physician communication before such technologybecomes widely distributed. Furthermore, we pro-pose a research agenda to provide data that is use-ful for eareful shaping of the communications infra-structure.

The Present State of Electronic Medicine

Internet technologies have become useful toolsfor medical practice. Online, physicians can searchthe medical literature and find both synoptie andfull-text medical journal content (6-10). Patientshave access to medical information, self-help andsupport groups, and even medical experts (11-13).The World Wide Web can be used to link patientdata across multiple institutions for retrieval by pro-viders at the point of service or by researchers (14,15). The Internet also allows visual contact. Forexample, the National Library of Medicine's Tele-

See related article on pp 441-449.

©1998 American College of Physicians-American Society of Internal Medicine 495

medicine Initiative enables parents of premature in-fants to videoconference from home with the neo-natal intensive care unit (16).

Although the Intemet community has always re-sisted national and international efforts to regulateits structure or content (17, 18), a motion is underway to impose an organizational structure and mon-itor the quality of the largely unchecked, unstructured,and unregulated volumes of medical ijiformationfound on the Web (19-21). Journals, professionalorganizations (19, 21, 22), and the federal govern-ment (23-27) have proposed monitoring medicalinformation on the Internet and regulating tele-medicine technologies. Early federal efforts alongthese lines include the Congressional Telecommuni-cations Act of 1996 (28) as well as the U.S. Foodand Drug Administration's proposed oversight oftelemedicine applications and clinical software (22).

Direct electronic communication hnkages be-tween physicians and patients are also evolving, al-though so far these are under considerably less scru-tiny. One such linkage is simple e-mail. Approximately40 000 000 persons in the United States use theInternet (29), and anecdotal reports are emerging ofits use for communication between physicians andpatients. In university health service settings, inwhich both patients and physicians tend to haveaccess to e-mail, an initial investigation has demon-strated the potential for widespread acceptance ofelectronic patienl-physician communications by thismethod (30, 31).

The Promise of E-Mail

E-mail can connect physicians with patients,thereby increasing access to care, enhancing patienteducation, augmenting screening programs, and im-proving adherence to treatment plans. Barriers toaccess often arise simply because physicians can bedifficult to reach (32). Unlike telephone conversa-tions, which require both parties to be available atthe same time, e-mail, like voice mail, is an asyn-chronous mode of communication (33, 34), essen-tially creating continuous access to the health caresystem.

A widening gap is developing between the crucialneed for transmitting more information and the rel-atively few and often brief face-to-face opportunitiesfor communication between physicians and patients.The quality of these personal encounters is furtherdiminished by the need for physicians to addressadministrative issues, such as referrals, insuranceapprovals, and rejected claims, during precious con-tact time. Inadequate communication, now more therule than the exception, leads to increased stress(35), diminished satisfaction (36, 37), decreased ad-

herence to therapeutic regimens (38, 39), and ele-vated risk for malpractice claims (40). Linking pa-tients and physicians through e-mail could increasethe involvement of patients in the supervision anddocumentation of their own health care, processesthat may activate patients and contribute to im-proved health (41-44). An example of electroniclinkages activating patients is CHESS (Comprehen-sive Health Enhancement Support System), an inter-active computer-based system used to supportpersons with AIDS and HIV infection (45).

Telephone and voice mail technologies havebeen effective in screening for mental disorders andsubstance abuse (46, 47). For example, computer-generated telephone reminders can improve compli-ance with preschool immunization visits (48). Stan-dard e-mail or e-mail with an interface allowingstructured data entry may allow more effective triageand automation of messaging than voice and videomail do.

Whether e-mail between physicians and patientscan have advantages for the therapeutic relationshipis a question worthy of investigation. We proposea research agenda structured around foreseeableproblem areas in electronic patient-physician com-munication.

Potentiai Pitfails of E-Maii and the Needfor a Research Agenda

Inappropriate Use of Communication Tools

Health care providers need a framework forchoosing the communication mode that is most ap-propriate for each situation. Certain kinds of com-munication needs may be satisfied through e-mail.For example, the patient may use e-mail to make anappointment. He or she may request general infor-mation, such as a list of low-sodium foods, or spe-cific information, such as a modified insulin dosagebased on home monitoring of glucose levels. Thephysician might initiate e-mail contact to conductroutine guidance and education (for example, toadvise a new mother to put her baby to sleep on hisback), to remind a patient of an upcoming visit, orto check on a patient's progress (for example, in asmoking cessation program). In contrast, the use ofe-mail might be contraindicated in some areas, andface-to-face or telephone contact might be required.Use of e-mail by patients for urgent needs couldlead to problems not being addressed quicklyenough. It might also be inappropriate for physi-cians to use e-mail to communicate abnormal orconfusing test results or to relay bad news. Thediagnosis of a new problem requiring a complex anddynamic dialogue might be best handled synchro-nously. Face-to-face contacts are optimal for making

496 15 September 1998 • Annals of Internal Medicine • Volume 129 • Number 6

many diagnoses, although elaborate telephone tri-age systems have been successfully used for thispurpose (49). Preference for the use of e-mail willdepend on the context as well as on the individualphysician and patient. Just as some physicians nowfeel more comfortable than others with telephonemedicine, so must physicians exercise personal dis-cretion in their use of e-mail.

Used properly, e-mail may promote increasedcontact between physicians and patients. By increas-ing opportunities for communication before and af-ter visits, e-mail might help optimize the value ofpersonal encounters. Empirical studies, as well asexpert consensus panels, are needed to develop gen-eral guidelines to direct patients and clinicians touse the contextually appropriate mode of communi-cation. Evidence-based indications and contraindica-tions for e-mail in the medical context must beclearly specified so that e-mail is used appropriatelyand does not become a barrier to telephone orface-fo-face contact.

Security and Confidentiality

The confidentiality of medical information (50)and the privacy of e-mail are paramount. Patients orphysicians who use e-mail in the workplace for med-ical interchange are not assured of confidentialityand may inadvertently expose sensitive details ofillness or social circumstances to an employer. Fur-thermore, patients who use family e-mail accountsat home may lack privacy from spouses, children, orparents. Medical account addresses eould be distinctfrom other personal or professional ones. Medicale-mail addresses and the messages generatedthrough them should be reliably documented in andlinked to the patient's medieal record. Such linkagecan be accomplished in various ways, from simplyincluding a patient identifier to embedding a hyper-text link to a Web-enabled medical record (14).

A critical decision will be the selection of a set ofnational health identifiers, as called for by theHealth Portability Act of 1996 (23). Naive use ofbroadly disseminated identifiers may be unwise. Forexample, one proposed identifier, the Social Secu-rity number, can easily be used to identity a patientand link his or her health record to other records,such as those in financial and marketing databases(51). The Computer Scienee and Telecommunica-tion Board of the National Research Council (52)reeommends proeedures to authenticate the user'sidentity and to encrypt stored or communicateddata to prevent unauthorized access. Perhaps thebest-developed technology for authenticating mes-sages is public key cryptography (53). This technol-ogy requires that every partieipant be assigned apair of "keys." One key is public and can be widelyshared over the Internet; the other key is private.

Messages encrypted with the public key can only bedecrypted by the private key and vice versa. Thepatient can send a message to his clinician by en-crypting it with the clinician's public key so thatonly the clinician can decrypt the message with herprivate key. Similarly, a elinician ean prove that sheis the sender of a message (that is, she can authen-ticate the message) by encrypting her name (orother identifier) with her private key. The patientcan then decrypt the clinician's name only with theclinician's public key. For public key cryptographyto be effective in the context of a health deliverysystem, research into rapid and secure methods fordistributing keys must be conducted. Methods bywhich consumers can easily apply such keys must bealso be developed.

Although electronic communications must beprotected from unauthorized interlopers, most vio-lations of the confidentiality of electronic data arecommitted by authenticated persons (52). To pre-vent these breaches within organizations, furtherstudy is needed to provide a framework for devel-oping institutional confidentiality policies, educationprograms, and effective legislation.

Guiding Patient Communication

A well-designed interface eould guide patients asthey create messages and could help triage mes-sages received at the physician's practice. Patientscould create messages by choosing from menu cat-egories; simple scheduling questions, for example,could be automatically directed to the front deskstaff, or prescription refill requests eould be directedto the practice nurse. The interface should also havesafety features to ensure appropriate communica-tions. For example, the patient could be required tospeeify the priority of a message. If the patientindicates that the message is urgent, the systemwould either display a warning screen suggesting atelephone call or a visit or trigger an alert to thephysician, such as an alarm or a page.

Many physicians are concerned that they eouldbe overwhelmed by long, numerous e-mail messagesfrom their patients. The interface could effect flowcontrol by tracking the number and length of mes-sages being sent by patients; when thresholds de-fined by the individual provider are exceeded, thesystem eouid suggest a telephone call or visit. Fur-thermore, the interface might passively encouragebrevity by offering only a small window for textentry and thereby suggesting the appropriate lengthof a message. Given the flexibility afforded by con-nectivity, the patient eould be given the option ofreceiving automated responses to certain routinequeries; this would remove the physician from cer-tain highly selected communication loops. Healthservices and informatics methods should be applied

15 September 1998 • Annals of Intemal Medicine • Volume 129 • Number 6 497

to study the effect of modifying the human-com-puter interface on communication. There is a fineline between filtering out inappropriate communica-tion and creating a barrier to access.

Hypotheses about the influence of e-mail oncosts and use of health services must be testedrigorously. Improved access to clinicians may reducethe need for some visits. Enhanced communicationabout health promotion and disease prevention pro-grams might reduce morbidity and mortality. On theother hand, e-mail communication and the formal-ization and documentation of the patient-physiciandialogue may increase both appropriate and inap-propriate use of services. Some patients may be-come nervous and seek urgent health care. Whenthe dialogue appears in the written record, somephysicians may be less willing to reassure patientsand may have a greater tendency to recommendhealth care visits or diagnostic testing.

Medicolegal LiabilityThe use of e-mail has complex medicolegal im-

plications. Most office staff will be unable to re-spond instantly to e-mail from patients. A safe andreliable system must somehow prevent the genera-lion of messages that require urgent or emergencyresponses if the messages might remain unansweredfor a long time. In addition, with the current pro-prietary e-mail systems, one party has no way ofknowing whether the other has received a criticalmessage. Again, research into the features of thehuman-computer interface is required. Electronicreceive-and-read receipts could show both cHnicianand patient that important contacts were made.

The inherently superior documentation of e-mailcommunication compared with telephone calls, forexample, could either protect against or increasephysician liability. The medical community shouldestablish guidelines for medical record storage andthe organization of e-mail correspondence. Oncee-mail is in widespread use between physicians andpatients, surveillance should be set up that will al-low the legal system to derive risk managementprinciples specific to the new and evolving situation.

Inequitable Access to TechnologyInequitable distribution of a new technology may

widen social disparities in health care access andoutcomes. An effective therapy that has differentialaccess according to socioeconomic characteristics orethnicities may cause a divergence in health statusoutcomes among segments of the population (54).Although Internet access is skewed toward wealth-ier, more educated users (55, 56), market forcesmay shape a substantially more equitable distribu-tion. Such businesses as telephone companies, cabletelevision providers, technology companies, and

mass media conglomerates have identified this vastmarket for information as a major area for invest-ment and development. Web-browsing technologywill soon be built into standard televisions. Medi-cine will need to adapt the technology, not create it.Penetration of this technology into the average U.S.home will probably parallel that of personal elec-tronic items other than high-end computers. In1995, only 33% of U.S. homes had personal com-puters, but 85% had videocassette recorders and97% had color televisions (57).

With increased use of the Internet for everydayconsumer commerce, repeated predictions of wide-spread degradation or "brownouts" in the quality ofservice have been made (58). The Next GenerationInternet, however, will provide a much larger trans-mission capacity (59). Even now, many techniquesare available to improve performance of the currentcapacity. For example, transmission of large files(such as a cardiac angiogram) can be done duringoff-peak hours.

Medical systems offering electronic medical com-munication should research and monitor rates ofaccess within their diverse patient populations.Cost-effectiveness analyses should establish themedical and economic advantages to the system ofpatient access to e-mail. The cost-effectiveness ofproviding technology to targeted patient populationsmight be studied. Even if widespread access isachieved, interface design research is still critical. Ifpatients arc required to deal with complex, poorlydesigned user interlaces or interfaces that require ahigh reading level whole segments of the popula-tion may be cut off from the benefits of the system.Voice and video capabilities of e-mail systems couldimprove access for patients who are not literate.

In addition, if widespread access is to be pro-moted, the technology must be used sensibly. Ratesof access will improve if software to communicatewith a health system or physician practice is acces-sible from any Internet-connected device. Con-versely, access will be diminished if patients arerequired to have specialized software on a localhard drive or a proprietary e-mail account (60).

Conclusions

E-mail between physicians and patients offerssubstantial promise as a way to improve access tohealth care, to let physicians reach out to patients,and to increase the involvement of patients in theirown care. Because electronic communication be-tween patients and providers has not yet occurredon a large scale, the opportunity still exists for pa-tients, physicians, and software engineers to evalu-ate the impact that the technology will have on

498 15 September 1998 • Annals of Iniemal Medicine • Volume 129 • Number 6

health care. The medical profession must ensurethat the tools used to link physicians and patientscan be used safely and appropriately by a multi-cultural, multihngual population with a wide rangeof literacy. New communication technologies mustnever replace the cmeial interpersonal contacts thatare the very basis of the patient-physician relation-ship. Rather, e-mail and other forms of electroniccommunications should be used to enhance suchcontacts. At this critical juncture, we have the op-portunity to proceed cautiously and to collect evi-dence to back up our assumptions about the opti-mal role of e-mail between physicians and patients.It would be useful to define explicit requirementsfor devices that will connect providers to patientsand to specify the indications for their use. Theform and function of electronic communicationtools and the choice of domains for their applica-tion will influence the nature of medical relation-ships and may ultimately affect the state of health inthe United States.

Acknowledgments: The authors thank Dr. Peter Szoliwits for in-spiring them to explore eleetronic physician-palient cummuniea-tion and Alison Clapp, librarian at Children's Hospital. Boston,Massachusetts, for help in conducting the literature review.

Requests for Reprints: Kenneth D. Mandl. MD. MPII, Division ofEmergency Medieinc, MA-OOl, Children's Hospital, 300 Long-wood Avenue. Boston, MA 02115.

Current Author Addresses: Dr. Mandl: Division of EmergencyMedicine. MA-OOl. Children's Hospital, 300 Longwood Avenue,Boston. MA 02115.Dr. Kohane: Division of Endocrinology, Children's Hospital. 300Longwood Avenue. Boston. MA 02115.Dr. Brandt: Departmeni of Social Medicine, Harvard MedicalSchool. 641 Huntington Avenue, Boston, MA 02115.

References

1. Fischer CS. America Calling: A Social History of the Telephone to 1940.Berkeley: UnJv CaltfornJa Pr; 1992:176,

2. Aronson SH. The Lancet on the telephone, 1876-1975. Med Hist 197721 •69-87

3. Rogers EM. Diffusion of Innovations. 4th ed. New York: Free Pr; 1995.4. Keen J. Rethinking NHS networking, BMJ, 1998;316:1291-3.5. Kane B, Sands DZ. Guidelines for the clinical use of electronic mail with

patients. The AMIA Internei Working Group, Task Force on Guidelines for theUse of Clinic-Patient Elearonic Mail. J Am Med Inform Assoc. 1998,5-104-n,

6. Clowniak JV. Medical resources on the Internet, Ann Intern Med 1995' 123-123-31,

7. Glowniak JV, Bushway MK. Computer networks as a medical resource.Accessing and using the Iniernet. JAMA. 1994:271:1934-9.

8. Chi-Lum Bl, Lundberg GD, Silberg W M . Physicians accessing the Internet,the PAI Project, An educational initiative |Editorial|. JAMA. 1996;275:1361-2,

9. Fraser HS, Kohane IS, Long WJ. Using the technology of the World WideWeb to manage clinical information. BMJ. 1997;314.1600-3.

10. Kassirer JR. Journals in bit5 and bytes. Electronic medical journals [Editorial],N EngI J Med, 1992;326:195-7.

11. Finding help online. Consumer Reports. 1997;62:27-31.12. Ferguson T. Health Online: How to Find Health Information, Support

Groups, and Self-Help Communities in Cyberspace. Reading, MA: Addison-Wesley; 1995,

13. Widman LE, Tong DA. Requests for medical advice from patients andfamilies to health care providers who publish on the World Wide Web. ArchIntern Med 1997,157:209-12.

14. Kohane IS, van Wingerde FJ, Fackler JC, Cimino C, Kilbridge P, Mur-phy S, et al. Sharing electronic medical records across multiple heteroge-neous and competing institutions, Proc AMIA Annu Fall Symp. 1996:608-12,

15. Halamka JD, Safran C. Virtual consolidation of Boston's Beth Israel and New

16.

17.

18.

19.

20,

21,

22.

23,

24.

25.26.

27.

28.

29,

30.

31.

32.

33.

34.

35.

36.

37.

38,

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

England Deaconess Hospitals via the World Wide Web Proc AMIA Annu FallSymp. 1997:349-53.Gray J, Jones PC, Phillips M, Gertman PC, Veroff D, Safran C Telematicsin the neonatal ICU and beyond: improving care for high-risk newborns andtheir families, Proc AMIA Annu Fall Symp, 1997:413-7.

. Diffie W, Landau S. Privacy on the Line: The Politics of Wiretapping andEncryption. Cambridge, MA. MIT Pr; 1998,

. Gurak L. Persuasion and Privacy in Cyberspace: The Online Protests overLotus Marketplace and the Clipper Chip, New Haven, CT: Yale Univ Pr; 1997,Jadad AR, Qagliardi A. Rating health information on the Internet, navigat-ing to knowledge or to Babel? JAMA. 1998;279:611-4.Peters R, Sikorski R. Navigating to knowledge. Tools for finding informationon the Internet. JAMA, 1997;277:505-6.Kassirer JP. The next transformation in the delivery of health care [Editorial].N EngI J Med. 1995;332.52-4.Miller RA, Gardner RM. Summary recommendations for responsible moni-toring and regulation of clinical software systems. American Medical Infor-matics Association, The Computer-Based Patient Record Institute, The MedicalLibrary Association, The Association of Academic Health 5cience Libraries, TheAmerican Health Information Management Association, and The AmericanNurses Association, Ann Intern Med. 1997;127 842-5,Health Insurance Portability and Accountability A a of 1996, Public Law No,104-191, Section 1173, USC 201, 1996.Weeda DF, O'Flaherty NF, Food and Drug Administration regulation ofblood bank software: the new regulatory landscape for blood establishmentsand their vendors. Transfusion. 1998:38:86-9.Wenner H. Software validation Medical Device Technology, 1997;8:22-5.FDA weighs policing electronic patient record. Hosp Case Manag. 1997;5'81-3,Gross TP, Ke55ler LG. Medical device vigilance at FDA, Stud Health TechnofInform. 1996;28:17-24.Jones MG. Telemedicine and the national information infrastructure: are therealities of health care being ignored? J Am Med Inform Assoc, 1997;4-399-412Hof RD, Browder S, Elstrom P. Special report: Internet communities. Busi-ness Week, 1997;5 May:64-92.Fridsma DB, Ford P, Altman R, A survey of patient access to electronicmail: attitudes, barriers, and opportunities [Abstract!, Proceedings—18th An-nual Symposium on Computer Applications in Medical Care, 5-9 November1994,Neill RA, Mainous AG 3d, Clark JR, Hagen MD. The utility of electronicmail as a medium for patient-physioan communication. Arch Fam Med, 1994;3:268-71.Young KT, Davis K, Schoen C, Parker S. Listening to parents. A national5urvey of parents with young children. Arch Pediatr Adolesc Med 1998 i52 '255-62Balas EA, Jaffrey F, Kuperman GJ, Boren SA, Brown GD, Pindroli F, etal. Electronic communication with patients. Evaluation of distance medicinetechnology. JAMA. 1997;278:1 52-9,Allely EB. Synchronous and asynchronous telemedicine, J Med Syst 1995'19:207-12.Roter DL, Hall JA, Kern DE, Barker LR, Cole KA, Roca RP. Improvingphysicians' interviewing skills and reducing patients' emotional distress. Arandomized clinical trial. Arch Intern Med. l995;155:1877-84.Bertakis KD. The communication of information from physician to patient: amethod for increasing patient retention and satisfaction. J Fam Pract, 1977;5:217-22

Greene MG, Adelman RD, Friedmann E, Charon R. Older patient satis-faction with communication dunng an initial medical encounter, Soc Sci Med1994:38:1279-88.Waitzkin H. Doctor-patient communication. Clinical implications of socialscientific research. JAMA. 1984;252:2441-6,Korsch BM. What do patients and parents want to know? What do theyneed to know? Pediatrics. 1984;74(5 Pt 2):917-9Levinson W, Roter DL, Mullooly JP, Dull VT, Frankel RM. Physician-patient communication. The relationship wilh malpractice claims among pri-mary care physicians and surgeons. JAMA. 1997;277553-9.Szolovits P, Doyle J, Long WJ, Kohane I, Pauker SG. Guardian Angel:Patient-Centered Health Information Systems, Report TR-604, Cambridge,MA. MIT Laboratory for Computer Science; 1994,

Dertouzos ML. What Will Be. How the New World of Information WillChange Our Lives. New York. Harper Collins; 1997.Slack WV, Hicks GP, Reed CE, Van Cura U. A computer-based medicaf-history system, N EngI I Med. 1966;274:194-8,Bysinger B. Using the Internet to support provider-patient collaboration.Health Manag Technol. 1997; 18.44-5,Gustafson DH, Hawkins RP, Boberg EW, Bricker E, Pingree S, Chan CL.The use and impact of a compuler-based support system for people [ivingwith AIDS and HIV infection, Proc Annu Symp Comput AppI Med Care1994:604-8.Baer L, Jacobs DG, Cukor P, O'Laughlen J, Coyle JT, Magruder KM.Automated telephone screening sunyey for depression. JAMA. 1995,273:1943-4,Kobak KA, Taylor LH, DottI SL, Greist JH, Jefferson JW, Burroughs D,et al. A computer-administered telephone interview to identify mental disor-ders, JAMA, 1997;278:905-10.Linkins RW, Dini EF, Watson G, Patriarca PA. A randomized trial of theeffectiveness of computer-generated telephone messages in increasing immu-nization visits among preschool children. Arch Pediatr Adolesc Med 1994-148:908-14.Poole SR, Schmitt BD, Carruth T, Peterson-5mith A, Slusarski M. After-

15 September 1998 • Annals of Intemal Medicine • Volume 129 • Number 6 499

hours telephone coverage: the application of an area-wide teiephone triageand advice system for pediatric practices. Pediatrics. 1993;92:670-9.

50. Rind DM, Kohane IS. Szolovits P, Safran C, Chueh HC. Bamett GO.Maintaining the confidentiaiity of medical records shared over the Internetand the Worid Wide Web. Ann intern Med. 1997; 127 138-41

51. Szolovits P, Kohane 1. Against simple universai health identifiers, J Am Medinform Assoc 1994;1.316-9.

52. National Research Council. Committee on Maintaining Privacy and Securityin Health Care Appiications of the National Information infrastruaure. For theRecord: Protecting Eiectronic Health information. Washington. DC: NationalAcademy Pr; 1997.

53. Rivest R, Shamir A, Adieman L A method for obtaining digitai signaturesand public-key cryptosystems. Communications of the Association for Com-puting Machinery. 1978;21:120-6.

54. Hamvas A, Wise PH. Yang RK, Wampler N5. Noguchi A. Maurer M M ,et al. The infiuence of the wider use of surfactant therapy on neonatai

mortality among blacks and whites. N Engi J Med. 1996:334:1635-40,55. Hoffman Dt, Novak TP. Bridging the racial divide on the internet. Science.

1998:280:390.56. Katz J, Aspden P. Motivations for and barriers to Intemet usage: results of

a national pubiic opinion survey. 1996. Avaiiabie at http://wvw/.mari<ie.org.57. Meeker M, DePuy C. The internet Report—Morgan Staniey. New York:

Harper Collins, 1996.58. Metcalfe B. Predicting the internet's catastrophic collapse and ghost sites

galore in 1995 infoworid. 1995;4 Dec:61.59. Ackerman M, Bailey D, DuBois R, Endicott 0, Fowier J, Hui B, et ai.

NexL Generation Internet initiative: NGi implementation Pian. Ariington, VA:National Coordination Office for Computing, Information, and Communica-tions; 1998. Available at h1tp://v/ww.ccic,gov/ngi/implementation/

60. Mandl KD, Katz SB, Kohane IS. Social equity and access to the World WideWeb and e-mail, impiications for design and impiementation of medical ap-plications. Prcc AMIA Annu Fall Symp. 1998; [In press).

When he got warm, he got happy again. Several years ago Dr. Hawkins, who treatedboth Hellman and Sylvia on the ridge and then in the hospital, told me that, if 1 wereburned and wanted to be as happy as Joe Sylvia had been, I should get terriblyburned. "Then," he said, "your sensory apparatus dumps into your bloodstream." Headded, "Usually it takes until the next day to clog your kidneys. In the meantime, itis possible to have spells when you think you are happy,"

Norman MacleanYoung Men and FireChicago: Univ of Chicago Pr; 1992

Submitted by:Malcolm L. Brigden. MDKelowna, British Columbia, Canada

Submissions from readers are welcomed. If the qucitalion is published, the sender's name will be acknowl-edged. Please include a complete citation, as done for ;iny reference—Tte Editor

500 15 September 1998 • Annals of Internal Medicine • Voktme 129 * Number 6