Email consultations in health care: 2—acceptability and safe applicationBMJ 2004; 329 doi: http://dx.doi.org/10.1136/bmj.329.7463.439 (Published 19 August 2004) Cite this as: BMJ 2004;329:439
- Josip Car, doctoral student in patient-doctor partnership1,
- Aziz Sheikh, professor of primary care research and development ()2
- 1 Department of Primary Care and Social Medicine, Imperial College London, London W6 8RP
- 2 Division of Community Health Sciences: GP Section, University of Edinburgh, Edinburgh EH8 9DX
- Correspondence to: A Sheikh
Email may have an important role in augmenting and facilitating communication between patients and healthcare professionals.1 In this article we summarise the evidence describing public and professional attitudes to using email in routine clinical care and explore issues relating to ensuring such use is safe. Specifically, we aim to answer three questions:
How acceptable is email consulting to members of the public and healthcare professionals?
How can the quality and safety of email consultations be ensured?
How can we safeguard patient information when using email for consulting?
Attitudes towards email consultations
Recent research suggests that members of the public increasingly want personalised care delivered by a range of approaches, the exact means depending on individual preference and the nature of the problem.2 w1 Several descriptive studies provide consistent evidence that people who use email would like email access to their doctors.3–7 w2-w8 For example, a recent US Harris poll revealed that 90% of survey respondents would welcome the opportunity to communicate with their doctors by email, with 56% of those surveyed stating that ability to communicate with their doctors electronically would influence their choice of doctor. Perhaps more surprising is that 37% indicated that they would be willing to pay for email access to their doctor.w9 This wish to communicate electronically seems to be driven by public perceptions that email communication offers the potential for quick and convenient access to healthcare advice that can supplement the traditional face to face consultation.3 w5 w9
Between 1% and 10% of the US public communicate with their doctors electronically, most in only a limited capacity.7 8 w4 w6 w9 Electronic communications that the public would particularly like to see available include follow up emails after visits to doctors, receipt of personalised medical information, obtaining test reports, and submission of charts for monitoring chronic conditions such as diabetes and asthma. The public also thinks that doctors should use automated systems to help patients better manage preventive care (such as email reminders for impending flu vaccination).w1 Many patients also wish to use email to book or cancel appointments, arrange non-urgent consultations, and request repeat prescriptions.5
Patients' satisfaction with email communication has been shown to be high, and this mode of communication is preferred over telephone calls by many for discussing non-urgent problems (see fig 1). However, satisfaction decreases sharply as response time increases, with patients expecting to receive a response within 48 hours of sending an email.5 7 w10 This is perhaps unsurprising because one of the main perceived benefits of email communication is its speed.w1
Most doctors in Western countries now have access to email at work (such as via the NHSNet in Britain). Surveys in the United States and Europe show that many doctors use email to communicate with patients (up to 20% in Europe and 25% in the United States) but that most do so with only a small proportion of their patientsw9 w11-w14 and generally only at the request of patients.w9 The main concerns of those who offer these services to patients are the protection of confidentiality, potential for errors and liability, and securing payment.6 w4 w11 w15 w16 Other concerns include identifying clinical situations where email communication is likely to be inefficient (perceived difficulties of responding to complicated messages, for example) and the challenges of incorporating email into existing work patterns without increasing overall workload and costs.6 w3 w4 w13 w17
Although doctors are selective in choosing which patients they communicate with by email, their selection criteria remain unclear.9 Qualitative studies of doctors communicating by email with their patients suggest that they view it as a useful addition to the communication options already available and believe that it offers the potential to enhance management of chronic diseases, facilitate patient education, improve continuity of care, and increase flexibility in responding to non-urgent issues.9 w10 w18 w19 There is, however, little evidence to support these views.1
In 1999, 45 US specialty and state medical societies joined forces to create the Medem Network to connect doctors and patients online.w20 This online resource encourages patients to let their doctors know that they are interested in email access to care if such services are not already locally available. This mode of consulting has received little attention in UK plans for modernising information technology in the NHS.w21
Safe application of email consultations
Wider adoption of email consultations seems to be hindered by concerns, voiced mainly by doctors, about the quality, security, and safety of email consultations. These concerns are receiving considerable attention as managed care organisations and insurance companies (mainly in the United States) strive to realise the potential that email consulting has to offer.
Email is generally viewed as a good means of communicating simple information and non-urgent requests between patients and doctors. In one study, however, up to 90% of patients who used email to communicate with their doctors relayed important and sensitive medical information electronically.3
In order to ensure safe and high quality email consultations, each specialty should develop policies and standards that will earn the trust of the public and healthcare professionals.10–13 w22 w23 These should include guidance on organising the service, training, and strategies for mitigating risk during all stages of processing emails (receipt, triage, and response).11 Selection of optimal software for email consultation is extremely important (see box 1).w24
A key concern is whether email management is appropriate? Whenever in doubt doctors should revert to safer modes of consulting. A second concern is that the key points of a consultation have been correctly understood. Doctors can facilitate understanding by following the principles used in face to face or telephone encounters—such as the use of simple language, encouraging patients to ask questions, and summarising the main points covered.2 14 Additionally, a feedback loop may be used, whereby patients report what action they will take. If the subject being addressed is likely to require several emails back and forth, it is best to advise the patient to have a face to face or telephone consultation (see fig 1).
It may (initially) be appropriate to use a standard protocol clearly delineating the types of email communications that will be considered (for example, appointment scheduling, reporting of home records such as peak expiratory flow or blood pressure, ordering repeat prescriptions, obtaining test results, and consultations for a predefined set of conditions). Unsuitable topics, because of their complexity or sensitivity (because of the associated security concerns), may also be predefined. Patients should be advised not to use email for urgent communications. Similarly, when a doctor wants to ask a patient about symptoms that may require prompt action (such as chest pain or shortness of breath) a synchronous mode of consulting should be used. It may also be appropriate that each email includes a reminder about the importance of alternative forms of communication for emergencies.
Patients and doctors should communicate only through designated email addresses and services. Triage nurses may screen emails, as they do telephone calls, before they are routed to the appropriate person for a response.15 An automatic reply from the clinic can acknowledge receipt of a patient's email, and patients should be requested to acknowledge reading a doctor's email. Emails should be flagged as “unresolved” until an acknowledgment is received. Standardising specific communications (use of customised templates or protocols) to meet the needs of various specialties and tasks (such as repeat prescriptions) may make communication easier and increase quality and safety (see fig 2).
Box 1: Key features of optimal software for email consultations
Ease of adoption (combining with existing technologies)
Adaptability to an organisation's unique requirements for managing personal health information
Seamless operating with existing infrastructures
Enabling communication over various operating systems (such as Windows or Linux) and software programs
User friendliness—easy to set up, manage, and use by doctors and patients
Effective, invisible security over wired and wireless environments (without users needing to be aware of safeguards)
Easy authentication methods
Integration with existing medical records systems
Possibility of the use of customised templates for email consultations
Automation functions (such as automatic replies)
System for preventing generation of messages to an addressee if previous messages remained unanswered for longer than set permissible time
Integrated customisable message content filtering (if desired)
Track and audit messaging system
Archiving and logging
Further considerations include which system is less error prone, needs less intensive support, and is more productive
The written record of email consultations enables close monitoring and evaluation of appropriateness and safety. Whereas face to face and telephone consultations are rarely recorded verbatim (typically being documented with only a few key words), email provides direct evidence of patient-doctor conversation. Thus, email consultations have the potential to facilitate accurate record keeping. However, if the system is not seamlessly integrated with medical records, as is the case with many healthcare organisations, quality of recording may be poor. At present, many doctors need to “copy and paste” email messages into records or print out and file a paper copy of them. Collaboration between developers of secure email software and providers of electronic health records is needed to achieve their seamless integration.
Currently, email consultations with unknown patients are considered unsafe, and there are no agreed standards for such consultations (see box 2). A pragmatic approach to an unsolicited email from an unknown person or someone unregistered for use of email consultations is to reply with a standard disclaimer.
Box 2: Email consultations with an unknown person
Some healthcare professionals consult by email with patients they have never met in person.16 17 w25 w26 For some patients, such anonymous email communications may be less intimidating than face to face consultations. These communications may help them understand the nature of their problem and decide to seek help.w27-w30 However, the quality of such advice may vary greatly, and patients cannot be sure that replies are actually written by a doctor.
Currently, determining the identity and credibility of “cyberdocs” is difficult. In future, this could be overcome—for example, by using digital signatures (electronic proof of identity, which could also confirm that a doctor is officially licensed to provide such a service). Until standards are established, email consultations outside pre-existing patient-doctor relationships should not be considered as a method of rendering medical care.10 11 16 17 w30-w32
Suggestion for an automatic reply to an unsolicited email “Your email has not been read by any of the staff from Wood Thorpe Medical Centre. This is an automatic reply from a computer to an unsolicited email. No medical advice can be given without prior informed consent for email consultations. If you need any clarification about this automatic computer reply, please telephone 0171 540 4987.”
Additional educational resources
Safeguarding patient information
The ethical considerations, professional etiquette, and legal rules that guide traditional communication between healthcare professionals and patients are equally applicable to email consultations.10 18 19 w33Patients should be informed of the potential risks and benefits of email use, the ramifications, safeguards for privacy and confidentiality, and the practice or hospital policies on when and how to use email.
Ideally, informed consent should be obtained from patients before email communication is started.w34 Patients should know who will process and have access to emails, including the times when the addressee is unavailable. They should be informed about the time in which email will be read and replied to—such as by the end of the next working day. An email should be forwarded (for example, to a specialist) or edited only with the sender's consent.
Email use in health care has developed without encryption. The security of unencrypted email is low, and email content can be inadvertently disclosed on the internet or local computer. Many countries now oblige healthcare organisations to follow the same strict data protection rules as do commercial institutions such as banks. As well as firewalls and conventional network security to protect content stored on an organisation's network, software that helps to achieve secure email communication (either web based or with standard email software) is available from several companies (such as Secure Data in Motion, Sigaba,w35 Medemw20). A critical factor in any solution for ensuring security is its user friendliness, and this may differ for patients and doctors and with different clinical settings and purposes.w36 Safeguarding patient information also depends equally on paying due attention to organisational and technical considerations. Doctors risk breaching patient confidentiality if they use non-secure email with patients.
The Institute of Medicine considers “a sustained partnership between patients and clinicians” as a critical element of primary care.w37 Email communication can help achieve this at a time when the personalised doctor-patient relationship is under threat from the increasing demands on health services.
Using email for patient-doctor communication increases patient choice in the way health care is received. To date, its use has largely been patient led, with healthcare organisations slow to adopt it.20 Making email communication more readily accepted as a part of routine medical practice should be a key objective of the UK NHS information technology strategy. Widespread use in clinical practice requires the coordinated action of health professional organisations, patient representative groups, policy developers, and the information technology industry.
National surveys show that patients increasingly want to be able to communicate with healthcare professionals by email
Few doctors currently do so: professional concerns centre on the quality of consultations, confidentiality, liability, and the challenge of recovering fees
Early email use in health care has grown without an adequate supporting infrastructure to address security issues
Ensuring privacy, confidentiality, and security of information is vital for email consultations, and various user friendly safeguards are now becoming available
Email consultations mark a radical shift from the traditional oral modes of communication: both patients and doctors need education in how to use them safely and effectively
Extra references (w1-w37) are listed on bmj.com
We thank Professors Azeem Majeed, George K Freeman, and Martyn R Partridge for their critical comments on an earlier draft of this manuscript.
Contributors AS and JC conceived the idea for this review. JC conducted the searches, evaluated the study quality, and analysed the data. AS contributed to the search design, quality evaluation, and data analysis and interpretation. JC wrote the first draft of the paper; both authors jointly wrote the paper subsequently. Both authors are guarantors for the paper.
Funding JC is supported by research awards from the Ministry of Education, Science and Sport, Slovenia, Ad Futura Foundation, and Universities UK (ORS award).
Competing interests None declared.