Email for the provision of information on disease prevention and health promotion

Abstract

Background

Email is a popular and commonly used method of communication, but its use in health care is not routine. Its application in health care has included the provision of information on disease prevention and health promotion, but the effects of using email in this way are not known. This review assesses the use of email for the provision of information on disease prevention and health promotion.

Objectives

To assess the effects of email for the provision of information on disease prevention and health promotion, compared to standard mail or usual care, on outcomes for healthcare professionals, patients and caregivers, and health services, including harms.

Search methods

We searched: the Cochrane Consumers and Communication Review Group Specialised Register (January 2010), the Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library, Issue 1 2010), MEDLINE (1950 to January 2010), EMBASE (1980 to January 2010), CINAHL (1982 to February 2010), ERIC (1965 to January 2010) and PsycINFO (1967 to January 2010). We searched grey literature: theses/dissertation repositories, trials registers and Google Scholar (searched July 2010). We used additional search methods: examining reference lists, contacting authors.

Selection criteria

Randomised controlled trials, quasi‐randomised trials, controlled before and after studies and interrupted time series studies examining interventions where email is used by healthcare professionals to provide information to patients on disease prevention and health promotion, and taking the form of 1) unsecured email 2) secure email or 3) web messaging. We considered healthcare professionals or associated administrative staff as participants originating the email communication, and patients and caregivers as participants receiving the email communication, in all settings. Email communication was one‐way from healthcare professionals or associated administrative staff originating the email communication, to patients or caregivers receiving the email communication.

Data collection and analysis

Two authors independently assessed the risk of bias of included studies and extracted data. We contacted study authors for additional information. We assessed risk of bias according to the Cochrane Handbook for Systematic Reviews of Interventions. For continuous outcome measures, we report effect sizes as mean differences (MDs). For dichotomous outcome measures, we report effect sizes as odds ratios (ORs). We conducted a meta‐analysis for one primary health service outcome, comparing email communication to standard mail, and report this result as an OR.

Main results

We included six randomised controlled trials involving 8372 people. All trials were judged to be at high risk of bias for at least one domain. Four trials compared email communication to standard mail and two compared email communication to usual care. For the primary health service outcome of uptake of preventive screening, there was no difference between email and standard mail (OR 0.93; 95% CI 0.69 to 1.24). For both comparisons (email versus standard mail and email versus usual care) there was no difference between the groups for patient or caregiver understanding and support. Results were inconclusive for patient or caregiver behaviours and actions. For email versus usual care only, there was no significant difference between groups for the primary outcome of patient health status and well‐being. No data were reported relating to healthcare professionals or harms.

Authors' conclusions

The evidence on the use of email for the provision of information on disease prevention and health promotion was weak, and therefore inadequate to inform clinical practice. The available trials mostly provide inconclusive, or no evidence for the outcomes of interest in this review. Future research needs to use high‐quality study designs that take advantage of the most recent developments in information technology, with consideration of the complexity of email as an intervention.

Plain language summary

Email used by health professionals to send patients/caregivers information on promoting health and preventing disease

Email is widely used in many sectors and lots of people use it in their day to day lives. The use of email in health care is not yet common. One use for it is for health professionals to send patients/caregivers information on how to be healthy and avoid disease. This review examines how patients, healthcare professionals and health services may be affected by using email in this way.

We found that there was not much evidence on the effects of using email to give people information on disease prevention and health promotion. We found only six trials with 8372 participants in total. All of the trials had elements of bias. Four studies compared email to standard mail as a method of communication, and found that using email instead of mail did not make any difference to patient or caregiver understanding, or the uptake of preventive screening. Two studies compared email with usual care, and found that using email instead of usual methods of information delivery did make any difference to patient or caregiver understanding and support, or patient health status and well‐being. We were unable to properly assess email's impact on patient or caregiver behaviours/actions as the results were mixed.

As there is a lack of good quality evidence for whether email can be used by healthcare professionals to provide information to patients or caregivers on how to stay healthy and avoid disease, we need to think about how to get good measurable information on this. Future studies should follow advice on good ways of carrying out and presenting research. It would be useful if they could look at the costs of using email and take into account ongoing changes in technology.

Summary of findings

Summary of findings for the main comparison. Summary of findings: Email communication compared with standard mail for preventive health purposes.

Patient or population: Adultsa Settings: Different healthcare settingsb Intervention: Email communication for preventive health purposesc Comparison: Standard mail
Outcomes	No of Participants (studies)	Quality of the evidence (GRADE)	Impact
Patient or caregiver behaviours/actions	96 (2 studies)	⊕⊝⊝⊝ very lowd,e,f,g,h	Evidence is inconclusive. Email was shown to make a significant difference to one measure of patient or caregiver behaviours/actions when compared to standard mail, but not shown to make a difference for seven measures of patient or caregiver behaviours/actions when compared to standard mail.
Patient or caregiver understanding and support	44 (1 study)	⊕⊝⊝⊝ very lowi, j,k	Email communication was not shown to make a significant difference to patient or caregiver understanding and support when compared to standard mail.
Health service outcome: uptake of preventive screening	1859 (3 studies)	⊕⊝⊝⊝ very lowl, m,o	Email communication was not shown to make a significant difference to uptake of preventive screening when compared to standard mail.
Healthcare professional outcomes	0 (0)	See impact	NOT MEASURED
Harms	0 (0)	See impact	NOT MEASURED
GRADE Working Group grades of evidence High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate.

^a Adults over 50 invited or due for colorectal cancer screening, cardiac rehabilitation patients, females aged between 40‐75 due for mammography. 
 ^b Outpatient settings, hospital owned health and fitness facility, Health maintenance organisation. 
 ^c Invite to pick up test kit, personalised reminder email, fitness tips email. 
 ^d One study has 4 of 6 domains at high risk of bias, the other has 5 of 6. 
 ^e Evidence is inconclusive with some measures showing a significant difference and others not. 
 ^f One study set in specific setting: health and fitness facility. Both studies have specific populations (cardiac rehabilitation patients, patients fifty or over). 
 ^g Confidence intervals visibly wide for several measures. Sample sizes modest. 
 ^h One study failed to report outcomes of one arm of a two arm study because of low survey response rates. 
 ⁱ Study has 5 of 6 domains at high risk of bias. 
 ^j Study set in health and fitness facility and has specific population of cardiac rehabilitation patients 
 ^k Confidence intervals are wide for all measures and sample size is modest. 
 ^l Three studies, one with 3 of 6 domains at high risk of bias, one with 4 of 6 domains at high risk of bias, another with 1 of 6 domains at high risk of bias and 3 at unclear risk. 
 ^m One study failed to report outcomes of one arm of a two arm study because of low survey response rates.

^o Studies set in specific settings (outpatient settings or Health maintenance organisation) with specific participants (adults over 50 invited or due for colorectal cancer screening and females aged between 40‐75 due for mammography).

Summary of findings 2. Summary of findings: Email communication compared with usual care for preventive health purposes.

Patient or population: Adultsa Settings: Different healthcare settingsb Intervention: Email communicationc Comparison: Usual care
Outcomes	No of Participants (studies)	Quality of the evidence (GRADE)	Impact
Patient or caregiver behaviours/actions	132 (2 studies)	⊕⊝⊝⊝ very lowd,e,f,g	Evidence is inconclusive. Email was shown to make a significant difference to two measures of patient or caregiver behaviours/actions when compared to usual care, but not shown to make a difference for five measures of patient behaviours/actions when compared to usual care.
Patient health status and well‐being	49 (1 study)	⊕⊝⊝⊝ very lowh,i,j	Email communication was not shown to make a significant difference to patient health status and well‐being when compared to usual care.
Patient or caregiver understanding and support	49 (1 study)	⊕⊝⊝⊝ very lowk,l,m,n	Evidence is inconclusive Email was not shown to make a significant difference to three measures of patient or caregiver understanding and support when compared to usual care. Three measures were not compared statistically so we cannot make any definitive conclusions about them.
Health service outcomes	0 (0)	See impact	NOT MEASURED
Healthcare professional outcomes	0 (0)	See impact	NOT MEASURED
Harms	0 (0)	See impact	NOT MEASURED
GRADE Working Group grades of evidence High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate.

^a Parents (caregivers) of children with chronic constipation/encopresis, patients who had lost weight >5% of initial body weight. 
 ^b Paediatric gastroenterology clinic, outpatient weight loss clinic. 
 ^c Reminder to pick up test kit, email reminder for log in to educational website, weight loss tip. 
 ^d One study has 2 of 6 domains at high risk of bias and 3 at unclear risk, the other has 3 of 6 at high risk. 
 ^e Evidence is inconclusive with some measures showing a significant difference and others not. 
 ^f Both studies set in specific settings (paediatric gastroenterology and weight loss clinic) with specific participants. 
 ^g Confidence intervals presented for one measure are wide,for other measures they are not presented. Two measures presented as median values as data not normally distributed. 
 ^h 3 of 6 domains at high risk of bias. 
 ⁱ Study set in an outpatient weight loss clinic with specific participants (patients who had lost weight >5% of initial body weight). 
 ^j Confidence intervals wide, two measures reported as median values. 
 ^k 3 of 6 domains at high risk of bias. 
 ^l Evidence inconclusive with estimates not available for some measures. 
 ^m Study set in an outpatient weight loss clinic with specific participants (patients who had lost weight >5% of initial body weight). 
 ⁿ No confidence intervals, all measures presented as median values.

Background

Related systematic reviews

This review forms part of a suite of reviews, incorporating four other reviews:

• email for communicating results of diagnostic medical investigations to patients (Meyer 2012);

• email for clinical communication between patients/caregivers and healthcare professionals (Atherton 2012b);

• email for clinical communication between healthcare professionals (Pappas 2012); and

• email for the coordination of healthcare appointments and attendance reminders (Atherton 2012a).

The use of email

Email is easy to use, widely available internationally and inexpensive. It is used in many areas of life, including banking, travel and retail. Despite the ubiquity of email in day‐to‐day life and in other sectors of the economy, its use in the healthcare sector is still not routine(Neville 2004; Dixon 2010) although it is increasing. Factors driving the trend of increasing email use include: the natural demographic shift towards an increasing proportion of people comfortable with using technology‐driven care solutions; and higher demands on healthcare resources with, for instance, the advent of increased chronic care and demand for more preventive screening, resulting in a focus on working more efficiently (OECD 2006).

Where email communication has been demonstrated in healthcare settings, its use has included requesting prescriptions, booking appointments and for clinical consultation (Kleiner 2002; Gaster 2003; Kittler 2003; Neville 2004; Castren 2005; Anand 2005).

Email for the provision of information on disease prevention and health promotion

Email can be used as a one‐way, healthcare professional‐to‐patient method of providing information on disease prevention and health promotion. Email contact from a healthcare professional may disseminate information more widely than conventional methods of providing information, reaching a wider audience or a specific target audience, such as older people.

Email can be used to improve the content and quality of communication with patients; for instance by sending patients emails with attachments containing tailored health information. Such information can increase knowledge and encourage positive healthcare choices, in turn leading to better health outcomes (Anderson 2003; Hesse 2005; Hardey 2008).

Disease prevention

Email can be used to send invitations for a service and also reminders, for instance for attendance at screening programmes. Traditionally such invitations and reminders have been administered via the post or telephone to inform patients when they are due for services such as paediatric immunisations, cervical smear tests, mammography and heart disease risk assessment (Car 2004a; Stone 2002).

Websites which provide a reminder service exist, but they are limited in number and based in the US. They offer to provide emailed reminders for various screening services such as mammography and diabetes testing at a pertinent time (College of American Pathologists 2008).

Email dissemination of information can be used to aid smoking cessation, counsel on contraception and advise on protection against sexually transmitted diseases, amongst other conditions (Virji 2006).

Health promotion

Email may provide a suitable method for disseminating health promotion messages to patients; bridging the gap between a need to share information with patients and the limited opportunity for face‐to‐face clinician to patient contact (Car 2004a). Such health information is now widely available in an online format, for example via the National Health Services (NHS) website 'NHS direct', or the UK National Asthma Campaign's e‐helpdesk (Car 2004a). This website provides information on asthma triggers, medicines and treatments and controlling asthma (Asthma 2008).

Despite the credibility of such websites, many others are less reliable. So that patients can avoid misleading or inaccurate data, various clinicians in the US are using web‐based information prescriptions; these are "prescriptions of specific, evidence‐based information to manage health problems" (D'Alessandro 2004). They allow healthcare professionals to direct patients and their families to high quality and appropriate information, which the patient can then browse at their leisure, and keep for reference. The prescriptions are regularly updated and can be arranged so as to be appealing and engaging (Ritterband 2005).

Email can add a time saving element to consultation; this is because information provided by the clinician via email, or from a website highlighted by a clinician, is much more in depth than that conveyed in a short consultation or a brochure. Patients can read such information in their own time, and keep it for reference (Anderson 2003). In a qualitative study, US physicians reported that they felt email was a useful educational tool and that patients appreciate being sent information by email to supplement the consultation (Patt 2003).

Advantages and disadvantages

The key advantages of email for the provision of information on disease prevention and health promotion include the following (adapted from Freed 2003; Car 2004a).

• Timely and low cost delivery of information (relative to conventional mail and provision of educational materials) (Houston 2003)

• The capacity to place hyperlinks in an email, leading to appropriate educational material.

• Read receipts can be used to confirm that communications have been received.

• Relative to verbal communication, the written nature of the communication can be of value as reference for the patient, aiding recall and, if desired, improving communication to other family members.

• Information in this format can be easily and inexpensively updated in accordance with new evidence; and can be customised for individual patient needs (Ritterband 2005).

• Patients are party to additional information about their health condition which they can review whenever and wherever they like, at their desired speed (Ritterband 2005).

• Email addresses usually stay constant when an address or telephone number changes (Virji 2006) making this reliable way of maintaining communication with transient patients.

• Easier communication method for patients with disabilities, and with patients that are temporarily overseas, such as seconded employees (Goodyear‐Smith 2005).

There are, however, some potential downsides such as the following:

• There is evidence of patient and physician concerns regarding privacy, confidentiality and the potential misuse of patient information (Harris 2001; Kleiner 2002; Moyer 2002; Katzen 2005).

• Physicians may be wary of the potential for email to generate an increased workload (Katz 2004; Podichetty 2004).

• Medico‐legal issues may exist (including around informed consent and use of non\‐encrypted email) (Bitter 2000).

• There is the potential to widen health inequalities via the digital divide. As new technologies replace old systems, it has been suggested that certain sectors of the population are being left behind with regard to access and use of these services, such as the elderly, non‐English speakers and those in lower income groups (Kleiner 2002; Katz 2004; Goodyear‐Smith 2005; Virji 2006).

• Technological issues may occur, such as recipients having a full mailbox causing email to bounce back to the sender (Virji 2006).

• Systems may be at risk of failures, such as a loss of the link to a central server (a computer which provides services used by other computers, such as email) (Car 2008a). There may be several causes for technological system failure; from local power failure to natural disasters.

• There is a potential for human error which can lead to unintended content or incorrect recipients.

Quality and safety issues

The main quality and safety issues around the use of email for disease prevention and health promotion have included: confidentiality; potential for errors and ensuing liability; identifying clinical situations where email consultation is inefficient or inappropriate; and incorporating email into existing work patterns (Kleiner 2002; Gaster 2003; Gordon 2003; Hobbs 2003; Houston 2003; Car 2004b).

Privacy and confidentiality are a formidable challenge in the adoption of email communication (Car 2004b; Katz 2004). Patients are more likely to use this type of communication if they have access to the Internet from home, rather than from work, because of privacy issues (Fridsma 1994). Family email accounts can mean a lack of privacy (Mandl 1998). Web messaging systems can address issues around security and liability that are associated with conventional email communication, since they offer encryption capability and access controls (Liederman 2003). However not all healthcare institutions are capable of providing such a facility, and instead rely on standard email (Car 2004b).

Medico‐legal issues are of substantial concern when implementing email communication in practice.These include potential liability for security breaches allowing a third party to access confidential medical information (Car 2004b). Suggestions for minimising the legal risks of using email in practice have included: adherence to the same strict data protection rules that must be followed in business and industry; and adequate infrastructure to provide encrypted, secure email transit and storage (Car 2004b).

Despite fears about increasing workloads, methods such as a web‐based information system for disseminating information are designed to be easy to use for both patient and practitioner; they can be updated quickly and easily, and information can be personalised using data about patient needs to generate suitable algorithms (D'Alessandro 2004; Ritterband 2005).

Patient opinion of such systems is also important, since distributing information is only useful if patients access and review it upon receipt. In a trial of a web‐based information prescription concerning childhood constipation, reasons given by parents for not accessing the recommended website included 'I forgot' and 'I didn't have time' (Ritterband 2005). Thus non‐compliance may prove to be an issue when using email for information provision.

These issues are wide ranging and encompass both healthcare professional and patient perspectives. We planned to identify all issues of quality and safety arising in the included studies.

Forms of electronic mail

In the absence of a standardised email communication infrastructure in the healthcare sector, email has been adopted in an ad‐hoc fashion and this has included the use of unsecured and secured email communication.

Standard unsecured email is email which is sent unencrypted. Secured email is encrypted; encryption transforms the text into an un‐interpretable format as it is transferred across the Internet. Encryption protects the confidentiality of the data, however both sender and recipient must have the appropriate software for encryption and decoding (TechWeb Network 2008).

Secure websites are distributed by secure web servers. Web servers store and disseminate web pages. Secure servers ensure data from an Internet browser is encrypted before being uploaded to the relevant website. This makes it difficult for the data to be intercepted and deciphered (TechWeb Network 2008).

There are significant differences in terms of these applications. Bespoke secure email programmes may incorporate special features such as the ability to show read receipts (in order to confirm the patient has received the correspondence) and, if necessary, facilities for receiving payment (Liederman 2005). However they are costly to set up and may require a greater degree of skill on the part of the user than standard unsecured email (Katz 2004). For the purpose of the review all methods are included although secured versus unsecured email would be considered in a subgroup analysis.

Methods of accessing email

Methods of accessing the Internet and thus an email account have changed with time. Traditionally, access would occur via a personal computer or laptop at home or work, connecting to the Internet using a fixed line. There are now several methods of accessing the Internet. Wireless networks (known colloquially as wifi) allow Internet connection to a personal computer, laptop computer or other device wherever a network is available (TechWeb Network 2008).

Internet connection is also possible via alternative networks using mobile devices. This includes access via mobile telephones to a wireless application protocol (WAP) network (rather than to the world wide web) or to third generation (3G) network. Adaptors connecting to a universal serial bus (USB) port can be used to access the 3G network using a laptop computer (TechWeb Network 2008). Therefore email can be accessed away from the office or home in a variety of ways.

For the purposes of this review we included all methods of email access.

Objectives

To assess the effects of using email for the provision of information on disease prevention and health promotion, compared to standard mail or usual care, on outcomes for healthcare professionals, patients and caregivers, and health services, including harms.

Methods

Criteria for considering studies for this review

Types of studies

We included randomised controlled trials (RCTs), quasi‐randomised trials, controlled before and after studies (CBA) with at least two intervention and two control sites, and interrupted time series (ITS) with at least three time points before and after the intervention.

Due to the practicalities of organisational change in a healthcare environment, it can be difficult to randomise studies and therefore we broadened our inclusion criteria to consider quasi‐randomised trials and CBAs. ITS studies are potentially valuable in assessing the ongoing merits of a new technology which may required a 'settling in' period. We included trials with individual and cluster randomisation, and relevant trials with economic evaluations.

Types of participants

We considered all healthcare professionals or associated administrative staff as participants originating the email communication and we considered patients or caregivers as participants receiving the email communication, regardless of age, gender and ethnicity. We also considered participants copied into the email communication. We included studies in all settings, i.e. primary care settings (services of primary health care), outpatient settings (outpatient clinics), community settings and hospital settings. We did not exclude studies according to the type of healthcare professional (e.g. surgeon, nurse, doctor, allied staff).

Email communication was one‐way, from healthcare professionals or associated administrative staff originating the email communication to patients or caregivers receiving the email communication.

Types of interventions

We included studies in which email is used by healthcare professionals for providing information to patients or caregivers on disease prevention and health promotion. We included interventions that use email for providing disease prevention and health promotion‐related information, such as smoking cessation information, immunisation information, public health education, and invitations or reminders for a preventive health check up.

We included interventions that used email in any of the following three forms:

1. Unsecured standard email to/from a standard email account.

2. Secure email which is encrypted in transit and sent to/from a standard email account with the appropriate encryption decoding software.

3. Web messaging, whereby the message is entered into a pro‐forma which is sent to a specific email account, the address of which is not available to the sender.

We included all methods of accessing email, include broadband via a fixed line, broadband via a wireless connection, connecting to the 3G network and connecting to the WAP network.

We excluded trials which considered the general use of email for healthcare professional‐patient contact for multiple purposes but did not separately consider health promotion or prevention information. Studies in which email was one part of a multifaceted intervention were included where the effects of the email component were individually reported, even if they did not represent the primary outcome. However these were only included where they achieved the appropriate statistical power. Where this could not be determined, or where it was not possible to separate the effects of the multifaceted intervention, we excluded these studies.

We also included invitations for routinely‐administered preventive screening, such as cervical screening and associated reminders and/or management. All other types of appointment and associated reminders were excluded from this review but considered in the parallel review on appointment and attendance reminders (Atherton 2012a). However, if an appointment had concerned both preventive and routine healthcare, such studies would have been included in both reviews.

We considered comparisons between outcomes of email communication and no intervention, as well as other modes of communication such as face‐to‐face, written material, postal letters, calls to a landline or mobile telephone, text messaging using a mobile telephone, and where applicable, automated versus personal emails.

Types of outcome measures

Primary outcomes of interest focused on whether the email has been understood and acted upon correctly by the recipient (patient or caregiver) as intended by the sender (healthcare professional or associated administrative staff), and secondary outcomes focused on whether email was an appropriate mode of communication.

Primary outcomes

Healthcare professional outcomes resulting from whether the email has been understood and acted upon correctly by the recipient (patient or caregiver) as intended by the sender (healthcare professional or associated administrative staff), e.g. professional knowledge and understanding, professional behaviour, action or performance.

Patient or caregiver outcomes associated with whether email has been understood and acted upon correctly by the recipient (patient or caregiver) as intended by the sender (healthcare professional or associated administrative staff), e.g. patient/caregiver understanding, patient health status and well‐being (e.g. lipid levels), skills acquisition, support, treatment outcomes, and patient/caregiver behaviours or actions (e.g. information seeking, smoking cessation, physical activity, weight management, nutrition and stress management).

Health service outcomes associated with whether email has been understood and acted upon correctly by the recipient (patient or caregiver) as intended by the sender (healthcare professional or associated administrative staff), e.g. uptake of preventive checks or screening.

Harms e.g. effects on safety or quality of care, breaches in privacy, technology failures.

Secondary outcomes

Professional, patient or caregiver outcomes associated with whether email was an appropriate mode of communication, e.g. knowledge and understanding, effects on professional‐patient or professional‐caregiver communication or relationship, evaluations of care (such as convenience, timeliness, acceptability, satisfaction).

Health service outcomes associated with whether email was an appropriate mode of communication, e.g. use of resources or time, costs, use of medical services, referrals, admissions.  

Search methods for identification of studies

Electronic searches

We searched the following electronic bibliographic databases.

• Cochrane Consumers and Communication Review Group Specialised Register (searched 8 January 2010)

• Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library Issue 1 2010)

• MEDLINE (OvidSP) (1950 to 5 January 2010)

• EMBASE (OvidSP) (1980 to 7 January 2010)

• PsycINFO (OvidSP) (1967 to 5 January 2010)

• CINAHL (EbscoHOST) (1982 to 2 February 2010)

• ERIC (CSA) (1965 to 7 January 2010)

We present detailed search strategies in Appendices 2 to 7. John Kis‐Rigo, Trials Search Coordinator for the Cochrane Consumers and Communication Group, compiled the strategies.

There were no language or date restrictions.

Searching other resources

Grey literature

We searched for grey literature, and ongoing and recently completed studies, in July 2010 using the following sources:

• Australasian Digital Theses Program (http://adt.caul.edu.au/)

• Networked Digital Library of Theses and Dissertations (http://www.ndltd.org)

• UMI ProQuest Digital Dissertations (http://wwwlib.umi.com/dissertations/)

• Index to Theses (http://www.theses.com/)(Great Britain and Ireland)

• Clinical trials register (Clinicaltrials.gov)

• WHO Clinical Trial Search Portal (www.who.int/trialsearch)

• Current Controlled Trials (www.controlled‐trials.com)

• Google Scholar (http://scholar.google.co.uk/) (we examined the first 500 hits)

We searched databases from their start date and there were no limitations by language. We kept detailed records of all the search strategies applied.

Reference lists

We examined the reference lists of retrieved relevant studies.

Correspondence

We contacted the authors of included studies across all five reviews (this review; Atherton 2012b; Atherton 2012a; Meyer 2012; Pappas 2012) for advice as to any further studies or unpublished data that they were aware of. Many of the authors of included studies were also experts in the field.

Data collection and analysis

Selection of studies

Two review authors (HA and PS) independently assessed the potential relevance of all titles and abstracts identified from electronic searches. We retrieved full text copies of all articles judged to be potentially relevant. Both HA and PS independently assessed these retrieved articles for inclusion. Where HA and PS could not reach consensus a third author, JC, examined these articles.

During a meeting of all review authors, we verified the final list of included and excluded studies. Where the description of a study was insufficiently detailed to allow the review authors to judge whether it met the inclusion criteria, we contacted study authors to obtain more detailed information to allow a final judgement regarding inclusion or exclusion. We retain detailed records of these communications.

Data extraction and management

HA and PS extracted data from included studies using a standard form derived from the data extraction template provided by the Cochrane Consumers and Communication Review Group. We extracted the following data:

• General information: Title, authors, source, publication status, date published, language, review author information, date reviewed.

• Details of study: Aim of intervention and study, study design, location and details of setting, methods of recruitment of participants, inclusion/exclusion criteria, ethical approval and informed consent, consumer involvement.

• Risk of bias: data extracted depended on study design (see Assessment of risk of bias in included studies).

• Participants: Description, geographical location, setting, number screened, number randomised, number completing the study, age, gender, ethnicity, socio‐economic grouping and other baseline characteristics, health problem, diagnosis, treatment.

• Health service: description, geographical location, setting, age, gender, population served, medical setting and clinical context of patients.

• Intervention: Description of the intervention and control including rationale for intervention versus the control (usual care). Delivery of the intervention including email type (standard unsecured email, secure email, web portal or hybrid). Type of clinical information communicated. Content of communication (e.g. text, image). Purpose of communication (e.g. providing information). Communication protocols in place. Who delivers the intervention (e.g. healthcare professional, administrative staff). How consumers of interventions are identified. Sender of first communication (health service, healthcare professional, associated administrative staff). Recipients of first communication (patient or caregiver). Any co‐interventions included. Duration of intervention. Quality of intervention. Follow up period and rationale for chosen period.

• Outcomes: principal and secondary outcomes e.g. desired behaviour change, methods for measuring outcomes, methods of follow‐up, tools used to measure outcomes, whether the outcome is validated.

• Results: for outcomes and timing of outcome assessment, control and intervention groups if applicable.

We piloted the data extraction template to allow for unforeseen variations in studies. Two review authors (HA and PS) independently extracted data from each included study. Any discrepancies between the review authors' data extraction sheets were discussed and resolved between the authors. When necessary, another review author (JC) was involved to resolve discrepancies.

Assessment of risk of bias in included studies

Two review authors (HA and PS) independently assessed the risk of bias of included studies, with any disagreements resolved by discussion and consensus, and by consulting a third author where necessary.

For RCTs (and quasi RCTs), we assessed and reported on the following elements that contribute to bias, according to the guidelines outlined in Higgins 2008:

• Sequence generation;

• Allocation concealment;

• Blinding (participants, personnel, outcomes assessors, data analysts);

• Intention‐to‐treat analysis;

• Incomplete outcome data;

• Selective outcome reporting.

We described the study and assigned a judgement relating to the risk of bias for each item. We used a template to guide the assessment of risk of bias, based upon the guidance by Higgins 2008, judging each item low risk of bias / unclear / high risk of bias. We summarised risk of bias for each outcome where this differed within studies.

We also assessed a range of other possible sources of bias and indicators of study quality, in accordance with the guidelines of the Cochrane Consumers and Communication Review Group (Ryan 2007), including:

• Baseline comparability of groups;

• Validation of outcome assessment tools;

• Reliability of outcome measures;

• Other possible sources of bias.

We presented the results of the risk of bias assessment in tables and incorporated the results of the assessment of risk of bias into the review through systematic narrative description and commentary about each of the quality items for each type of included study. This led to an overall assessment of the risk of bias across the included studies and a judgement about the possible effects of bias on the effect sizes of the included studies.

We contacted study authors for additional information about the included studies, or for clarification of the study methods as required.

Measures of treatment effect

For dichotomous data, we reported the odds ratio (OR) / risk ratio (RR) and 95% confidence intervals. For continuous data, we reported the mean difference (MD) and 95% confidence intervals. Where data were only available as median values we present them as such.

Assessment of heterogeneity

In the first instance we looked for heterogeneity via visual inspection of forest plots: if confidence intervals for individual studies have poor overlap it generally indicates the presence of statistical heterogeneity. Then we used a Chi² test to formally test for the presence of statistical heterogeneity. Where a meta‐analysis includes studies with a small sample size or where studies are few in number the Chi² test has low power. To allow for this, we used a P value of 0.10 (rather than 0.05) to determine statistical significance. As well as carrying out a Chi² test, an I² statistic was used. The test assesses the impact of heterogeneity on the meta‐analysis, rather than simply testing whether heterogeneity is present. The I² statistic quantifies inconsistency across the studies. It describes the % of the variability in effect estimates that is due to heterogeneity rather than sampling error.

We took the level of statistical heterogeneity into account when choosing the method of analysis for the review.

Assessment of reporting biases

Where data in the review have been standardised, we used a pooled funnel plot to check for publication bias.The funnel plot was produced using Review Manager 5 software.

Data synthesis

In a meeting of three of the review authors (HA, PS, JC) the included studies were assessed to determine whether they should be included in a quantitative meta‐analysis. We decided that we could only combine data for one outcome. This was because most outcomes were addressed by only one study. Two outcomes were addressed by two studies, but the data could not be pooled in a meta‐analysis owing to differences in the measures used to assess the outcome.

Where data were pooled, the choice of model was influenced by the level of statistical heterogeneity identified using both the Chi² and I² test. As the Chi² test was not significant and the value for the I² was low, indicating low levels of possible heterogeneity, we used a fixed‐effect model. A fixed‐effect meta‐analysis assumes that each study is estimating exactly the same quantity and that any variation between the results of the studies is due to chance. It more precise than a random‐effects model, because in the presence of statistical heterogeneity it usually has narrower confidence intervals. A random‐effects meta‐analysis assumes that the studies are not all estimating the same intervention effect. It can be used to incorporate heterogeneity among studies.The analysis was conducted in line with the guidance in the Cochrane Handbook of Systematic Reviews (Higgins 2008).

We applied the GRADE approach to assessing the quality of outcomes to each individual outcome, and produced a Summary of Findings table to outline the overall result for each outcome. In order to rate each outcome according to quality two authors (HA and PS) each independently rated the outcomes according to the five factors using guidance from the Cochrane Handbook and the guidance provide by the GRADE working group (Higgins 2008; GRADE 2010). Where ratings differed these were discussed until consensus was reached. Where consensus could not be reached a third author, JC, was consulted. We entered the finalised ratings into the GRADEpro software.

A typical Summary of Findings table produced in GRADEpro software contains a list of all important outcomes (usually primary outcomes per the review), a measure of the typical burden of these outcomes, the absolute and relative magnitude of effect (either/or), the number of participants and studies addressing these outcomes and a GRADE score for the overall quality of evidence for each outcome (rather than by study).  For the purposes of this review we adapted the table to account for the lack of data pooling. Despite the lack of numerical data the Summary of Findings table was a useful tool in summarising the findings of the review for the reader, and allowing for the quality of the outcomes to be assessed using the GRADE quality of the evidence framework.

The table was designed to contain the following:

• Each primary outcome (patient outcomes, health professional outcomes, health service outcomes and harms).

• Corresponding number of participants and studies.

• Quality of the evidence (GRADE score).

• Impact (via brief narrative summary).

We used an impact statement for each outcome to summarise the evidence available in the absence of statistical pooling. This statement was based on the measures of effect as per the study reports. Where the outcome had not been measured by any study in the review we stated this in the Summary of Findings table.

Subgroup analysis and investigation of heterogeneity

As it was only possible to perform a meta‐analysis for one outcome, which featured two studies, we were unable to conduct any statistical subgroup analysis. The methods that would have applied had we been able to combine data are outlined in Appendix 1, and will be applied to future updates of the review.

Sensitivity analysis

As it was only possible to perform a meta‐analysis for one outcome, which featured two studies, we were unable to conduct any sensitivity analysis. The methods that we would have applied had we been able to combine data are outlined in Appendix 1, and will be applied to future updates of the review.

Consumer input

We asked two consumers, a health services researcher (UK) and healthcare consultant (Saudi Arabia) to comment on the five completed email reviews before submitting them for the peer‐review process, with a view to improving the applicability of the review to potential users. The review also received feedback from two consumer referees as part of the Cochrane Consumers and Communication Review Group's standard editorial process.

Results

Description of studies

See Characteristics of included studies; Characteristics of excluded studies; Characteristics of ongoing studies.

Results of the search

As this review was one in a set of five looking at varying uses of email in healthcare, we conducted a common search for all five reviews (Pappas 2012; Atherton 2012b; Atherton 2012a; Meyer 2012). Relevant articles were allocated to the appropriate review after being assessed at the full text stage. For this review, we identified nine articles, reporting six studies that met the inclusion criteria (Chan 2008; Chaudhry 2007; Muller 2010; Pearson 2004; Ritterband 2005; Thomas 2008).

Figure 1 illustrates how the six included studies were selected.

1.

Flow diagram.

Included studies

We describe the main features of each of the six included studies in the Characteristics of included studies table.

Design

All of the included studies were randomised controlled trials. However not all authors described their studies as such. In one trial, participants were first stratified by private or public access arm depending on their email and Internet access, then both arms were randomised into control or intervention groups (Chan 2008). In another trial, participants were first assigned into either the intervention or control arms and then the intervention arm alone was randomised into a further subgroup consisting of email and standard mail groups (Pearson 2004). In another trial, participants were randomised into intervention and control arms and then a subgroup of Mayo Clinic employees in the intervention arm were further randomised into email or standard mail reminder groups (Chaudhry 2007). In one study, although participants were randomised into one of three arms, the study authors matched each arm for age and gender (Muller 2010).

Sample sizes

Two out of the six studies used power calculations (Muller 2010; Thomas 2008). The remaining four studies did not calculate sample sizes.

Setting

All six studies were conducted in high‐income countries (one in primary care, four in secondary or tertiary care and one in a Health Maintenance Organisation).

Five studies were conducted in the USA (Chan 2008; Chaudhry 2007; Muller 2010; Pearson 2004; Ritterband 2005) and one in the UK (Thomas 2008).

Primary care

For the study set in primary care, the participants were patients of the Division of Primary Care Internal Medicine at the Mayo Clinic in Rochester, Minnesota (Chaudhry 2007).

Secondary or tertiary care

Pearson 2004 was set in a hospital‐owned health and fitness facility serving as the regional medical centre for a 13‐county area of South California. Chan 2008, Ritterband 2005 and Thomas 2008 were all set in outpatient clinics. Thomas 2008 recruited participants from an outpatient weight loss clinic and in Chan 2008 participants were eligible only if they had attended the outpatient general internal medicine clinic at the University of Texas‐Houston for at least a year. Ritterband 2005 was set in a paediatric gastroenterology clinic.

Other care

The remaining study was conducted in a Health Maintenance Organisation in northwest United States with 18,847 enrolled patients (Muller 2010) but of these only 1,397 were randomised.

Participants

Participants in all studies were adults. However for one study, the adults were the families (caregivers) of the children being seen for the first time at a paediatric gastroenterology clinic with a chief complaint of chronic constipation and/or encopresis (Ritterband 2005). The intervention and outcomes were aimed at the families.

In Chan 2008, participants were men and women aged 50 years or over, and had to have attended an outpatient clinic. Similarly, Muller 2010 included men and women who were aged between 50 and 80 years. One study of breast cancer screening included women aged between 40 and 75 years of age (Chaudhry 2007). In Thomas 2008, those enrolled had already successfully completed a specific weight loss programme, where they initially had a Body Mass Index (BMI) greater than 30 and had then since lost at least 5% of their initial body weight. Participants enrolled in Pearson 2004 were part of a cardiac rehabilitation programme.

Sample sizes ranged from n = 55 (Thomas 2008) to n = 6665 (Chaudhry 2007).

Access to email

Participants in all studies had to have some form of access to email.

In Chan 2008, participants had to have private access to email and the Internet, or have an interest in public access. In one study (Pearson 2004), participants who self‐identified as 'online' computer users/owners were assigned to the intervention arm and then randomised into either the email and or standard mail group, and those remaining who were self‐identified as computer users, but were not 'online' users were assigned to the usual care arm. In Ritterband 2005, families had to have access to the Internet in their home and have an active email account. In Thomas 2008, participants were asked if they had access to email. In Chaudhry 2007 the employees randomised in the intervention group had to have a Mayo Clinic employee email account. Men and women enrolled in Muller 2010 had to have accounts on the Health Maintenance Organisation's secure email system.

Interventions

Purpose

Studies used email:

• to provide information on disease prevention (three studies: Chan 2008; Chaudhry 2007; Muller 2010),

• to provide health promotion information by delivering tips (two studies: Pearson 2004; Thomas 2008), and

• to promote an educational website (one study: Ritterband 2005).

The three studies using email to provide information on disease prevention used email reminders for routinely administered preventive screening. One study looked at mammography screening (Chaudhry 2007) and two looked at colorectal cancer screening (Chan 2008; Muller 2010). In addition, Chaudhry 2007 used other preventive screening tests and interventions including Papanicolaou smear, colorectal screening, pneumonia vaccine, influenza vaccine, tetanus vaccine and lipid screening. However in this study, patients due for mammography were randomised into an intervention and control group, then a subset of those in the intervention group (Mayo Clinic employees only) were further randomly assigned into either an email or standard mail group. For the purposes of this review, we only report data for this particular subset of participants meeting the review's inclusion criteria (email versus standard mail).

Of the two studies using email to provide health promotion information, one study delivered fitness tips as part of a maintenance cardiac rehabilitation programme (Pearson 2004) and another delivered weekly tips giving dietary, behavioural and exercise advice for weight maintenance (Thomas 2008).

Ritterband 2005 was the only study using email to promote an educational website. This specific website was prescribed by physicians to parents of children suffering from chronic constipation and/or encopresis.

The comparator arm for four studies was standard mail (Chan 2008; Chaudhry 2007; Muller 2010; Pearson 2004). In two studies the comparator was usual care (Ritterband 2005; Thomas 2008).

Two studies were multi‐interventional with three arms which included usual care, information delivered via email and standard mail (Muller 2010; Pearson 2004). However, the email versus standard mail was chosen as the comparator for these two studies because it was agreed that this comparison would tell us the most about email communication as an intervention.

The sender of the emails varied between studies with three studies using administrative staff (Chaudhry 2007; Muller 2010; Ritterband 2005), one study using a dietician (Thomas 2008), another using physicians (Chan 2008), and one using study authors who were also healthcare professionals (Pearson 2004).

Email form

Half of the included studies used secure email (Chan 2008; Chaudhry 2007; Muller 2010).

Of these, Chan 2008 used a secure health network email system (MyCareLINK). To view the 'NetLET' (InterNET LETter) the participant had to log onto MyCareLINK and cut and paste the uniform resource locator (URL) for the NetLET website into the browser window. Those who did not know how to use a computer were given a tutorial within two weeks of recruitment (Chan 2008).

Three studies did not specify whether the email form was unsecured or secured (Pearson 2004; Ritterband 2005; Thomas 2008). No studies used web messaging.

Email content and frequency

Single email reminders were used in two studies (Muller 2010; Ritterband 2005) and a third study included a website address in the email reminder (Ritterband 2005). The content in three studies constituted reminders for preventive screening (Chan 2008; Chaudhry 2007; Muller 2010). For the two studies delivering tips, these were both sent weekly (Pearson 2004; Thomas 2008).

However, additional information was provided in the personalised emails sent to participants in Chan 2008 which also included previous screening test results and a link to a secure web page with more information about colon cancer screening (CRCS). If no response was received within a week, the intervention (NetLET) was emailed to participants a second and third time at 1‐week intervals (Chan 2008). Another screening study used reminders, specifically two email reminders sent a month apart if there was no response after the first email reminder. If there was still no response after this, then participants received a scripted telephone call from an appointment secretary (Chaudhry 2007).

For the studies which sent weekly tips, one study was only for a total of six fitness tips for a six‐week duration containing principles of exercise relating to participation in cardiac rehabilitation (Pearson 2004) and the other study sent weekly tips for six months (Thomas 2008) containing dietary, behavioural and exercise advice. In the latter study, responses to the tips was not encouraged but participants were also sent a monthly email requesting a report of their current weight which required a response.

Outcomes

Healthcare professional outcomes

No study reported primary healthcare professional outcomes.

Patient/caregiver outcomes

Four studies reported primary patient outcomes (Chan 2008; Pearson 2004; Ritterband 2005; Thomas 2008):

• All four of these studies assessed outcomes specific to assessing patient behaviours or actions (Chan 2008; Pearson 2004; Ritterband 2005; Thomas 2008).

• Two of these studies assessed patient understanding and support (Pearson 2004; Thomas 2008).

• One of these studies assessed patient health status and well‐being (Thomas 2008).

Health service outcomes

Three studies reported primary health service outcomes (Chan 2008; Chaudhry 2007; Muller 2010), specifically the uptake of preventive screening.

Harms

No outcomes relating to harms were reported.

Excluded studies

We excluded two studies as they were not administered in a healthcare setting (Burgard 2006; Liguori 2007)(see Characteristics of excluded studies table).

Risk of bias in included studies

All of the studies had some degree of bias, most notably in the areas of allocation concealment and blinding (see Characteristics of included studies table). We sought further details from authors to allow us to adequately assess the risk of bias in the included studies where information was unclear. We were unable to contact the authors of one study (Muller 2010). We rated risk of bias elements as 'unclear' where authors did not respond, or where the information reported was insufficient to make a decision. Figure 2 summarises the risk of bias for each included study and Figure 3 summarises the risk of bias in included studies.

2.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

3.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Allocation

Three RCTs reported adequate sequence generation procedures (Chan 2008; Chaudhry 2007; Thomas 2008) and the procedures used for the other three included studies were unclear (Muller 2010; Pearson 2004; Ritterband 2005). In addition, only Thomas 2008 reported adequate allocation concealment using consecutively‐numbered sealed opaque envelopes. The remaining five studies lacked sufficient information for us to be able to judge the adequacy of allocation concealment.

Blinding

In all of the included studies, there was some form of inadequate reporting of blinding of participants and providers, but blinding may have been inappropriate due to the nature of the intervention which involved a form of communication with their healthcare providers and where participants had to actively take part, i.e. check their emails and in some studies respond back.

Thomas 2008 mentioned that blinding of the researcher and participants was not possible for the intervention group, as participants had to sign a consent form to confirm that they had been made aware of the insecure nature of electronic mail. Two studies made some attempt to blind physicians to group allocation (Chaudhry 2007; Ritterband 2005). No other study attempted to blind the researchers.

Incomplete outcome data

Incomplete outcome data were adequately addressed in three studies (Chan 2008; Muller 2010; Ritterband 2005). However for two of these studies, this was only the case for one of the reported outcomes (Chan 2008; Ritterband 2005).

Incomplete outcome data were inadequately addressed in one study (Pearson 2004), where although the numbers and reasons are given for the drop‐outs, it is not specified if they are from the email and/or standard mail groups of the intervention arm.

It was difficult to assess the adequacy of data reporting in one study as it was unclear whether an intention‐to‐treat analysis was carried out (Chaudhry 2007). The report states that some women declined to consent after randomisation and it is subsequently unclear what happened to these participants in the analysis.

Selective reporting

Published protocols were unavailable for the included studies; however it was possible to assess selective reporting based on a comparison of the outcomes specified in the methods sections and those reported in the results sections of the studies.

Two studies were free of selective outcome reporting (Muller 2010; Thomas 2008) while the remaining four studies showed evidence of selective outcome reporting. For example, in Pearson 2004, values were not reported where significance tests were carried out, and in Ritterband 2005, the results for an outcome are presented for the whole sample and not disaggregated by group.

Other potential sources of bias

Five studies were assessed as having a high risk of other sources of bias (Chan 2008; Chaudhry 2007; Pearson 2004; Ritterband 2005; Thomas 2008) and one study had an unclear risk of other sources of bias, due to missing information (Muller 2010).

The studies with a high risk of other sources of bias included recall bias (Chan 2008; Pearson 2004; Thomas 2008) and selection bias (Chan 2008; Chaudhry 2007; Pearson 2004; Ritterband 2005; Thomas 2008) amongst other sources. Only two studies mentioned using any sort of validated measures (Pearson 2004; Thomas 2008).

Effects of interventions

See: Table 1; Table 2

At Table 1 and Table 2 we present a summary of the results of the primary outcome measures. Data tables for individual outcomes can be found at Data and analyses. We present results organised by the two comparisons: email compared to standard mail, and email compared to usual care.

Email as method of communication compared to standard mail

Primary outcomes

Healthcare professional outcomes

None of the included studies reported primary healthcare professional outcomes.

Patient or caregiver outcomes

Patient or caregiver behaviours/actions

Two studies measured patient behaviours/actions (Chan 2008; Pearson 2004). Email largely did not make a difference to patient or caregiver behaviours/actions; there was a significant difference between groups for only one measure of the eight in this outcome category. In Chan 2008, participants in the email group were more likely to communicate with the doctor via email (OR 32.20 (95% CI 1.78 to 583.52)) (Analysis 1.1). There was no significant difference between groups for seeking information from the doctor by email (Analysis 1.2; Analysis 1.3; Analysis 1.4), or for discussing CRCS with the doctor, or other people (Analysis 1.5; Analysis 1.6; Analysis 1.7). In Pearson 2004, there was no significant difference between groups with regard to attendance at a cardiac rehabilitation programme (MD ‐0.30% (95% CI ‐10.95 to 10.35; Analysis 1.8)).

1.1. Analysis.

Comparison 1 Email versus standard mail: Primary outcome, Patient or caregiver behaviours/actions, Outcome 1 During study, communicated with doctor by email (private access arm only).

1.2. Analysis.

Comparison 1 Email versus standard mail: Primary outcome, Patient or caregiver behaviours/actions, Outcome 2 Information requested from doctor by email about a disease or condition (private access arm only).

1.3. Analysis.

Comparison 1 Email versus standard mail: Primary outcome, Patient or caregiver behaviours/actions, Outcome 3 Information requested from doctor by email about appointments, test results or prescriptions (private access arm only).

1.4. Analysis.

Comparison 1 Email versus standard mail: Primary outcome, Patient or caregiver behaviours/actions, Outcome 4 Information requested from doctor by email about 'other' topic (private access arm only).

1.5. Analysis.

Comparison 1 Email versus standard mail: Primary outcome, Patient or caregiver behaviours/actions, Outcome 5 Discussed CRCS with no‐one (private access arm only).

1.6. Analysis.

Comparison 1 Email versus standard mail: Primary outcome, Patient or caregiver behaviours/actions, Outcome 6 Discussed CRCS with doctor (private access arm only).

1.7. Analysis.

Comparison 1 Email versus standard mail: Primary outcome, Patient or caregiver behaviours/actions, Outcome 7 Discussed CRCS with 'other' (private access arm only).

1.8. Analysis.

Comparison 1 Email versus standard mail: Primary outcome, Patient or caregiver behaviours/actions, Outcome 8 % Programme attendance.

Patient or caregiver understanding and support

Overall, email communication was not shown to make a significant difference to patient or caregiver understanding and support when compared to standard mail. Only one study measured patient or caregiver understanding and support (Pearson 2004) using twelve measures, however there was no significant difference between email and standard mail groups with regard to perception of support in the cardiac rehabilitation programme. The measures were on three scales; four subscale measures and one overall support index measuring perception of functional support (Data and analyses 2.1 to 2.5); two subscale measures for perception of support specifically related to exercise participation each measured for family, friends and others (Data and analyses 2.6 to 2.10); and two measures on a Likert‐type scale measuring participatory intention scores (Analysis 2.11; Analysis 2.12).

2.11. Analysis.

Comparison 2 Email versus standard mail: Primary outcome, Patient or caregiver understanding and support, Outcome 11 Likert‐type scale: participatory intention scores: 6 months.

2.12. Analysis.

Comparison 2 Email versus standard mail: Primary outcome, Patient or caregiver understanding and support, Outcome 12 Likert‐type scale: participatory intention scores: 12 months.

Health service outcomes

Uptake of preventive screening

Email communication was not shown to make a significant difference to uptake of preventive screening when compared to standard mail. It was possible to pool data from two studies (Chan 2008; Muller 2010) for one measure (number of participants having colorectal screening). In order do this, the two different arms in Chan 2008 (public access arm and private access arm) were combined with Muller 2010. The meta‐analysis found no significant differences between the email and standard mail groups (OR 0.93 (95% CI 0.69 to 1.24)) (Analysis 3.1; Figure 4).

3.1. Analysis.

Comparison 3 Email versus standard mail: Primary outcome, Health service, uptake of preventive screening, Outcome 1 No. of participants having colorectal screening (completing either a FOBT, sigmoidoscopy or colonoscopy).

4.

Forest plot of comparison: 3 Email compared to standard mail: Health service outcome, uptake of preventive screening, outcome: 3.1 No. of participants having colorectal screening (completing either a FOBT, sigmoidoscopy or colonoscopy).

However, it must be noted that Chan 2008 assessed the number of participants returning a fecal occult blood test (FOBT) whereas Muller 2010 assessed a positive response to the intervention defined as either a FOBT, sigmoidoscopy or colonoscopy within three months of the study. The funnel plot for the two studies combined shows asymmetry (see Figure 5), indicating that perhaps smaller studies without statistically significant results remain unpublished. However the inclusion of only two studies, both with high/unclear risk of bias for some domains may be contributing to this asymmetry.

5.

Funnel plot of comparison: 3 Email versus standard mail: Primary outcome, Health service, uptake of preventive screening, outcome: 3.1 No. of participants having colorectal screening (completing either a FOBT, sigmoidoscopy or colonoscopy).

Uptake of preventive screening and immunisations including colorectal screening, Papanicolaou smear, influenza vaccine, tetanus vaccine, lipid screening and annual mammography were measured in one study with female participants only (Chaudhry 2007). No significant differences were found between the email and standard mail groups (Data and analyses 3.2 to 3.7).

Harms

No outcomes relating to harms were reported in any of the studies.

Secondary outcomes

There were no secondary outcomes in any of the studies.

Email as a method of communication compared to usual care

Primary outcomes

Healthcare professional outcomes

There were no healthcare professional outcomes reported in any of the included studies.

Patient or caregiver outcomes

Patient or caregiver behaviours/actions

The results for the measures of behaviours/action were equivocal; email was shown to make a significant difference to two measures of patient or caregiver behaviours/actions when compared to usual care, but not shown to make a difference for five measures of patient behaviours/actions when compared to usual care (Ritterband 2005; Thomas 2008).

In Ritterband 2005, parents (caregivers) in the email group had a significantly higher number of visits to a prescribed website than those in the usual care group (OR 2.99 (95% CI 1.17 to 7.65) (Analysis 4.1).

4.1. Analysis.

Comparison 4 Email versus usual care: Primary outcome, patient or caregiver behaviours/actions, Outcome 1 No. of visits to prescribed website within 1 week of clinic visit.

In Thomas 2008 there was no significant difference between groups for eating breakfast daily, drinking alcohol in last month, eating low fat food regularly and eating low sugar items regularly (Analysis 4.2 to Analysis 4.5), nor for fruit and vegetable consumption and exercise episodes per week, although these measures were only available as median values (Analysis 4.6; Analysis 4.7).

4.2. Analysis.

Comparison 4 Email versus usual care: Primary outcome, patient or caregiver behaviours/actions, Outcome 2 Eating breakfast daily.

4.5. Analysis.

Comparison 4 Email versus usual care: Primary outcome, patient or caregiver behaviours/actions, Outcome 5 Eating low sugar items regularly.

4.6. Analysis.

Comparison 4 Email versus usual care: Primary outcome, patient or caregiver behaviours/actions, Outcome 6 Median fruit and vegetable intake (portion number).

Median fruit and vegetable intake (portion number)
Study	Email (n=25)	Usual care (n=24)	P value for comparison
Thomas 2008	5	1.7	0.098

4.7. Analysis.

Comparison 4 Email versus usual care: Primary outcome, patient or caregiver behaviours/actions, Outcome 7 Median exercise episodes/week.

Median exercise episodes/week
Study	Email (n=25)	Usual care (n=24)	P value for comparison
Thomas 2008	3	0.5	0.13

Patient health status and well‐being

Email communication was not shown to make a significant difference to patient health status and well‐being when compared to usual care, with only one study measuring patient health status and well‐being (Thomas 2008) using six measures.

There was no difference between the groups for any of the measures. Participants were no more likely to be below their weight maintenance band (OR 4.38 (95% CI 0.45 to 42.39)), in their weight maintenance band (OR 1.50 (95% CI 0.48 to 4.65) or above their weight maintenance band (OR 0.37 (95% CI 0.11 to 1.26)) in the email group than the usual care group (Data and analyses 5.1 to 5.3). There was no significant difference in mean percentage weight loss between the email and usual care groups (MD 2.8 (95% CI 0.24 to 5.36; Analysis 5.4)). The final two measures were only available as median values. The authors report no significant difference between email and usual care groups for median body weight maintenance at 6 months (Analysis 5.5) and median amount of weight loss maintained at 6 months (Analysis 5.6).

5.4. Analysis.

Comparison 5 Email versus usual care: Primary outcome, patient health status and well‐being, Outcome 4 Mean % weight loss maintained at 6 months.

5.5. Analysis.

Comparison 5 Email versus usual care: Primary outcome, patient health status and well‐being, Outcome 5 Median body weight (kg) maintained at 6 months (IQR).

Median body weight (kg) maintained at 6 months (IQR)
Study	Email (n=25)	Usual care (n=24)	P value for comparison
Thomas 2008	94.6 (IQR: 38.5)	97.1 (IQR: 29.5)	Reported by author as 'not significant'

5.6. Analysis.

Comparison 5 Email versus usual care: Primary outcome, patient health status and well‐being, Outcome 6 Median amount of weight loss (kg) maintained at 6 months (IQR).

Median amount of weight loss (kg) maintained at 6 months (IQR)
Study	Email (n=25)	Usual care (n=24)	P value for comparison
Thomas 2008	9.6 (IQR: 10.9)	7.8 (IQR: 5.9)	0.12

Patient or caregiver understanding and support

Email was not shown to make a significant difference to patient or caregiver understanding and support when compared to usual care. Overall, only one study measured patient understanding and support (Thomas 2008).

The measures were only available as median values and the study reported that there were no significant differences reported in the study between groups for perceived effort of controlling diet (Analysis 6.1), undertaking activity (Analysis 6.2) and maintaining weight (Analysis 6.3).

6.1. Analysis.

Comparison 6 Email versus usual care: Primary outcome, patient or caregiver understanding and support, Outcome 1 Median perceived effort taken to control diet (IQR).

Median perceived effort taken to control diet (IQR)
Study	Email (n=24)	Usual care (n=25)	P value for comparison
Thomas 2008	‐1 (IQR: 2)	‐1 (IQR: 1)	Reported by author as 'non significant'

6.2. Analysis.

Comparison 6 Email versus usual care: Primary outcome, patient or caregiver understanding and support, Outcome 2 Median perceived effort taken to undertake activity (IQR).

Median perceived effort taken to undertake activity (IQR)
Study	Email (n=24)	Usual care (n=25)	P value for comparison
Thomas 2008	1 (IQR: 2)	1 (IQR: 3)	Reported by author as 'non significant'

6.3. Analysis.

Comparison 6 Email versus usual care: Primary outcome, patient or caregiver understanding and support, Outcome 3 Median perceived effort taken to maintain weight (IQR).

Median perceived effort taken to maintain weight (IQR)
Study	Email (n=24)	Usual care (n=25)	P value for comparison
Thomas 2008	0 (IQR: 2)	0 (IQR: 2)	Reported by author as 'non significant'

Health service outcomes

There were no health service outcomes reported in any of the included studies for this comparison. Muller 2010 was multi‐interventional with three arms which included usual care, information delivered via email and standard mail. We utilised the email versus standard mail data for this study because it was agreed that this comparison would tell us the most about email communication as an intervention. The results in this outcome category are thus reported under the 'Email versus standard mail' category.

Harms

There were no harms reported in any of the studies.

Secondary outcomes

There were no secondary outcomes reported in any of the studies.

Discussion

Summary of main results

Based on the findings of this review, it is not possible to determine whether email is of benefit in providing information on health promotion and disease prevention to patients and caregivers. The nature of the evidence base means that we are uncertain about the majority of the outcomes.

This review contains relatively few studies and these are of low quality. They provide mostly inconclusive, or no evidence for the outcomes of interest in this review. However, for a few primary outcomes there were no significant differences between groups.

The primary outcomes of interest were related to whether the email had been understood and acted upon correctly by the recipient as intended by the sender. We found that email made no difference to patient or caregiver understanding and support when compared with standard mail, and that evidence was inconclusive for patient or caregiver behaviours/actions (Table 1). For health service outcomes, email showed no difference in achieving uptake of preventive screening compared to standard mail (Table 1).

For the comparison of email to usual care, evidence was inconclusive for patient or caregiver behaviours/actions and patient or caregiver understanding and support. For patient health status and well‐being email made no significant difference when compared to usual care (Table 2).

No primary healthcare professional outcomes or harm outcomes were measured in any of the included studies falling under either comparison. No health service outcomes were reported under the comparison of email to usual care. There were no secondary outcomes reported in any of the included studies.

Overall completeness and applicability of evidence

There were no healthcare professional outcomes or harm outcomes reported by studies included in this review. The lack of healthcare professional outcomes is reflected in the lack of focus on healthcare professional understanding and behaviours, although these elements are addressed for patients and caregivers. This may reflect a perception that these outcomes are not important, as the recipient of the intervention is the patient or caregiver. The lack of outcomes relating to harm is apparent in the absence of information on privacy and security, technology failures and medico‐legal issues. We had expected to see these issues addressed in the included studies.

Two comparisons were identified; email versus standard mail and email versus usual care. Both comparisons could be said to mirror any potential real world use of email, as it would likely be introduced either to replace existing methods of communicating of such information (e.g. mail) or in addition to usual care.

All of the trials were carried out in high income countries. Five of the included studies were set in the United States, and the remaining trial in the UK (Thomas 2008). All of the studies in the review were published in English. English is the predominant language of both countries, and culturally and ethnically these countries are similar, each having largely white populations. However the UK and US differ with regard to health systems. The UK has a health system that offers universal coverage. The US does not, instead having a more mixed system with both government and insurance‐based coverage schemes. A significant number of people are not covered by these schemes. These characteristics may impact on the transferability of the results of the studies to other countries.

There is a perception that email as a distance technology might be beneficial for use in rural populations where attending healthcare settings may be more difficult for the patient (Hilty 2006). To date this has not been explored in a trial. One study was conducted in a university‐linked health setting (Ritterband 2005) and another in the Mayo Clinic which is a medical practice and medical research group with a medical school (Chaudhry 2007). Such settings provide a specific type of context for a research study. In the case of Chaudhry 2007, the participants were Mayo Clinic employees, and are therefore a very selected population. The results for a sample like this are unlikely to be generalisable to a wider setting. The trial is described as a feasibility study which may account for the choice of population; the authors may have chosen convenience over representativeness for the purposes of testing the feasibility of the intervention.

Only one of the included studies measured the ethnicity of participants (Pearson 2004) along with their education level and employment status. None of the other included studies measured the ethnicity or socio‐economic status (or any proxy measure such as education) of participants. This is surprising as these factors might be expected to have a significant impact on how patients use email. Pearson 2004 had a mostly Caucasian population (95%) and the majority of participants were retired or semi‐retired. In addition Pearson 2004 used being online (online computer user/owner) as a way to split the ‘intervention’ and ‘control’ groups, before randomising the ‘intervention’ group into email and standard mail groups. The email and standard mail groups were therefore comprised of self‐declared online computer owners/users.

Another study (Chan 2008) included two separate groups within which participants were randomised into email and standard mail groups. The two groups were those with private access (home and/or work) to the Internet and public access (willing to use Internet access via public library system). As we might expect according to the ‘digital divide’ those people who only had public access to the Internet were older, of non‐white ethnicities, had less formal education and lower incomes. This highlights a potential area for further investigation.

Muller 2010, a study set in the US, included only participants who had 12 months or more health insurance coverage with no more than one 45 day break before entering the study.  This excludes people who for whatever reason may not have had coverage.  Those people who are uninsured (and not eligible for the Medicaid health programme (US DHHS 2011) are more likely to be poor and have lower incomes (US DHHS 2005). Whilst this decision was likely made for methodological reasons, it should be noted when considering the generalisability of the study results.

Where demographic information was collected for these studies, it tended to concern age and gender. There was no mention of potential generational effects and no subgroup analysis based on characteristics such as age. If we had sufficient data, we would have conducted such an analysis.

We included various forms of email in this review: unsecured standard email, secure email and web messaging. Half of the studies in this review used secure mail and the other half did not specify email type. This made it difficult to consider the results of the studies according to email type. When the use of such technologies in healthcare is at an early stage, including all types of electronic mail together in one comparison can be justified, but future reviews may wish to consider the differences between the types of email and method of access, even if subtle. This could be in the form of a subgroup analysis where data can be combined.

Quality of the evidence

We have seen that the results of this review are equivocal and when interpreting the results we must also consider the high risk of bias in the included studies. A high risk of bias was reported for at least two domains in each study, with many domains remaining unclear.

We used the GRADE system (Guyatt 2008) to examine the quality of the evidence for each outcome, but as we were only able to combine data in a meta‐analysis for one health service outcome (uptake of preventive screening), the ratings should be seen as a guide to quality and strength of evidence, and not as definitive. The GRADE score for the outcomes in this review was 'very low quality'. This finding reiterates that we must view the results of this review with caution.

The heterogeneity in clinical settings and details of individual interventions, as well as the lack of data meant that we were only able to carry out a meta‐analysis for one measure of the health service outcome (uptake of preventive screening) for email versus standard mail. All other outcomes were represented by measures from one study only, where different measures were used to measure the same outcome within one study.

There was a high risk of performance and detection bias in these studies (blinding), with five out of six studies at high risk (Chan 2008; Chaudhry 2007; Muller 2010; Pearson 2004, Thomas 2008). This may be due to the nature of email; participants will always be aware that they are in the intervention group if they have consented to taking part in a study and then receive an email. Similarly those sending the email could not be blinded to study allocation and this was often the researcher conducting the study, or the patients’ physician. This lack of blinding may introduce bias into the studies.  Additionally, some of the risk of bias domains remained unclear, in some cases even after author contact (Chaudhry 2007; Ritterband 2005).  Authors for one study did not respond to attempted efforts to obtain further information regarding the risk of bias (Muller 2010). This makes it difficult to adequately assess the effects of interventions.

The number of participants in the individual studies varied, from n = 55 to n = 6665. The number of participants for the individual measures of the outcomes assessed in the review ranged from n = 44 to n = 1012. Worldwide, this is not a large number, especially when five of six studies originate from the same country.  Owing to the issues surrounding the quality of the evidence, we cannot be sure whether any consistencies in the results are genuine, for instance where all data presented for an outcome show no difference between the email group and the comparator. 

The search for this review was conducted in January 2010. A long period of time has elapsed between the search date and the publication of this review and this is a limitation as it is possible that relevant studies have been published in the interim period. To counter this, the review will be updated in the near future.

Potential biases in the review process

As well as database searches, we carried out an extensive search of the grey literature; this was helpful in providing a fuller picture of the evidence base.  As this is a fledgling field which has only become relevant alongside the increase in email use in day to day life, we can be certain to have searched the relevant time frame. By searching trial registers we identified several ongoing studies (Catz NCT00923624; Morgan ACTRN1260900092524; Samuelsson NCT01032265; Scrol NCT01077388).

Terminology is an ongoing problem with searching for evidence on new technologies, especially those used for communication. Several different terms can be used to describe email: electronic mail, electronic messaging, web messaging, web consultation, amongst many others. We were prepared for this in our search, choosing a wide selection of terms and using truncation of terms to ensure that all variations were found. It is possible that we missed other relevant terms. The changing nature of terms for technology should be considered in any future update of this review.

The broad criteria used in this review for types of studies, participants, interventions, and outcome measures will have ensured that studies were not excluded based on these factors. The two excluded studies in this review took place outside of healthcare settings (Burgard 2006; Liguori 2007). We chose to group the studies broadly and according to comparison, taking a pragmatic approach. This was so that we were able to get an overall picture of providing disease prevention and health promotion information by email. However the settings, participants and in some cases interventions were different between studies. For example, if we look at participants, some were suffering from a particular condition (Pearson 2004) and others were part of a general patient population (Chaudhry 2007).

We were only able to produce one funnel plot and this showed asymmetry, indicating that perhaps smaller studies without statistically significant results remain unpublished. Despite our sensitive search strategy, it is possible that there are data that were unavailable to us. For instance, if commercial companies have carried out trials and found these results to be negative or equivocal, they may choose not to publicise these results. The need for trial registration may not be apparent to organisations embarking on their first trials and doing so for commercial reasons. 

Agreements and disagreements with other studies or reviews

Other reviews and studies have considered the use of email for providing disease prevention and health promotion information.

Atherton 2010 reviewed the literature using systematic methods and did not restrict by study type. They included one of the same studies as in this review (Ritterband 2005) and this was one of two RCTs identified in their review. The remaining studies were a mixture of systematic reviews, questionnaire surveys and discussion articles. Atherton et al found that discussion of the use of email for communicating health promotion and disease prevention information was often speculative rather than evidence‐based. They identified useful aspects of the included articles. One of these was the potential for email to be used to send patients links to websites and another was the issue of equitable access to email. However, they concluded that the evidence base is not well established.  Atherton's review had a much wider remit than our review.

The two systematic reviews included in Atherton 2010 concerned the wider use of email in healthcare. McGeady 2007 reviewed the literature on the ‘impact of patient–physician web messaging on healthcare service provision.’ This non‐systematic review was restricted only by language, searching only for studies in English. It discussed the impact of email on disease management/prevention and how email might facilitate this, rather than using email directly for provision of information. Similarly Car 2004a systematically carried out a narrative review which described the potential for email in preventive healthcare with regard to sending invitations and reminders for preventive screening and assessment. The authors describe how it is not currently offered (in 2004) and state that there is no evidence on effectiveness or maximising impact.

Our review examined one‐way communication between healthcare professionals and patients or caregivers. The linked review ‘email for clinical communication between healthcare professionals and patients’ (Atherton 2012b) examined studies in which email was used for two‐way clinical communication between an individual healthcare professional and patient.  Whilst the studies in Atherton 2012b did not examine the content of the email messages between professional and patient, it is likely that patients may have requested information related to disease prevention and health promotion, or healthcare professionals may have utilised email to send such information to individual patients. This assertion is supported by the conclusions of the other reviews described here (McGeady 2007; Car 2004a), and by content analyses which show that requests for information are common (Sittig 2003; Stiles 2007; Byrne 2009) when communicating via email. This highlights the varied potential uses of email for the provision of this type of information beyond the one‐way use investigated in this review.

Until now, various studies have ‘set the scene’ for the use of email for preventive purposes.  This review differs in considering closely the quality of relevant studies and the evidence they provide specifically for the one way use of email for disease prevention and health promotion.

Authors' conclusions

Implications for practice.

Owing to the inconclusive evidence presented for the outcomes in this review, it is not possible to make recommendations for practice relating to the use of email for disease prevention and health promotion. However it should be noted that there was no evidence of harms. This review highlights an evidence gap in this area of research due to the lack of high‐quality evidence.

Implications for research.

Taking account of rapidly changing technologies

Although the evidence base is limited, we can assume that the use of email for disease prevention and health promotion will only grow.  Companies are increasingly providing services to facilitate the sending of email en masse.  Such companies offer publishing platforms for designing emails (style and content), sending emails in a targeted fashion, and tracking their receipt (MailChimp 2011; AWeber 2011; GetResponse 2011). These systems can be utilised to send information to many patients at once. It is also possible to target specific groups of patients based on information in their medical records; for example, patients with hypertension or women due for a cervical smear.  These systems are developing alongside a more general increase in the use of email in healthcare for multiple purposes.  In deciding where to focus future research, changes in technology like this should be considered. Rapid changes in technology make the outcomes from older studies difficult to interpret.  Presently, there is much greater penetration and use of email. Email now generally appears in html format rather than plain text, thus offering additional functionality, and web‐based technologies have also advanced, allowing email to be supplemented with links to websites, online video and audio, and social networks. Presuming that future changes in technology will follow a similar trajectory, we should be concerned with choosing outcomes that remain applicable in the face of such changes. This may involve concentrating on those elements that make email different from other methods of communication (asynchronous nature, stability of email address versus other personal details and the ability to store information) rather than making the intervention system itself the focus. Such factors do not change as rapidly with time as the technology changes. Otherwise randomised controlled trials may find their intervention dated by the time of their completion.  Qualitative research methods could be utilised to identify the important outcomes for the public, patients, physicians and other stakeholders so that trials identify relevant outcomes.

Assessing cost‐effectiveness

Only one of the included trials considered cost (Thomas 2008) by presenting cost of the intervention per patient, but this was not part of the trial and the intervention was compared to usual care and so no comparison was made. The costs of using email have reduced considerably in recent years.  Furthermore, the scalable nature of email (the costs of communicating with larger numbers of patients do not increase as rapidly as with other methods of communication) may mean it is more cost‐effective where it appears to be an equally effective method of communication.  The costs of using email may be the deciding factor even when all outcomes are positive. Reporting the costs of interventions and their comparison alongside the results of a trial would add context to the studies as the use of a successful email system may be prohibited on cost alone. Policymakers may wish to know whether email is more expensive or cost‐effective than other methods, as well as how effective it is. 

Standards for trial reporting and addressing complex interventions

In addition to considering the type of research that should be carried out in future, it is crucial to address the reporting of trials. Much of the uncertainty concerning the included studies in this review could have been avoided if standards for the planning, execution and presentation of trials were adhered to.  Use of the CONSORT statement (transparent reporting of trials) for RCTs (Schulz 2010) should be strongly encouraged, especially in the reporting of studies concerning complex interventions. The use of theoretical frameworks in evaluating complex interventions would be a valuable addition to any future research. The UK Medical Research Council guidance on developing and evaluating complex interventions (in trials or otherwise) is an example of such a framework (Craig 2008). The complexity of interventions such as email can make trial reporting in traditional journals with strict word limits difficult. Interventions may require much explanation and methods of analysis may be detailed. Newer online journals often offer the opportunity to place more detail in the appendix section of a publication and this is very useful for those wishing to read about a trial in full. 

The importance of trial registration

Registration of trials via online repositories such as clinicaltrials.gov should be strongly encouraged to discourage publication bias and selective outcome reporting. The lack of trial registration in our included studies may be due to those carrying out trials on communication systems within their own practices not seeing the need for registration as being as pressing as for drug trials and clinical interventions, or simply because the process of doing so is unfamiliar.

It is likely that future versions of this review and others like it will change as the evidence base expands and as the use of email becomes more common in health care.

Appendices

Appendix 1. Methods for application in future updates

Unit of analysis issues  

Unit of analysis issues did not arise in this review, however for the purposes of any future update we would address such issues in the following way:

Unit of analysis issues include those that may arise from the inclusion of cluster‐randomised trials, repeated measurements and studies with more than two treatment groups. If applicable the data would be analysed according to recommendations in the Cochrane Collaboration Open Learning Module on issues related to the unit of analysis (Alderson 2002).

Dealing with missing data  

None of the included studies had missing data. The methods that would have applied had we been required to obtain missing data are outlined below, and will be applied to future updates of the review:

If data were missing from the relevant comparisons we would attempt to contact the authors of the studies to obtain the information. If the authors could not be reached, or if the studies were found to be unsatisfactory on the basis of data provided, these studies would be excluded.

Subgroup analysis and investigation of heterogeneity  

The methods that would have applied had we had sufficient data are outlined below, and will be applied to future updates of the review.

If there had been sufficient data and it was appropriate in the context of the study, we would have examined the effect of certain studies on the pooled effects of the intervention.

1. Age

Consideration of the acceptability to different age groups (for both healthcare professionals and patients). This is important as there is clear evidence that the use of email is predicted by age with a clear tailing off in the generation who have not grown up in the digital age. It is therefore important to consider the intervention's effects in the groups which are accustomed to the technology, since it is likely to become more generalisable to the population as it ages. This would have been considered where the primary studies sought to consider age group from the outset. We would have distributed patients into three age subgroups: 0 to 17, 18 to 64, over 65. The choice of distribution was made on the basis of two surveys by The Pew Internet & American Life survey (Pew 2005).

2. Location

Location of the studies would also have been considered, since differing environments may condition the accessibility of the technology. For instance we would expect communication technologies and their accessibility to differ according to country and/or region within a country, such as rural or urban areas.

3. Type of email communication

Additionally we proposed to analyse the results by method of electronic mail utilised e.g. standard email versus a secure web messaging service.

4. Year of Publication

Lastly we would have considered results by year of publication, as those more recent studies may be more relevant given evidence of increasing usage and therefore assumed acceptability.

Other subgroups that may be considered where relevant include patients versus caregivers and socio‐economic status (where these data have been collected by the authors).

Sensitivity analysis  

The methods that we would have applied had we had sufficient data are outlined below, and will be applied to future updates of the review.

Studies deemed to be of lower quality after examination of individual study characteristics would be removed from the analysis to examine the effect on the pooled effects of the intervention. We would also consider the assessment of the risk of bias of included studies, as described above.

We would exclude studies according to the following filters:

• Outlying studies after initial analysis.

• Largest studies.

• Unpublished studies.

• Language of publication.

• Source of funding (e.g. public versus industry).

Other possible considerations for sensitivity analysis would include different measures of effect size (risk difference, odds ratios).

Appendix 2. MEDLINE (OvidSP) search strategy

1. computer communication networks/

2. limit 1 to yr="1996 ‐ 2002"

3. electronic mail/

4. (electronic mail* or email* or e‐mail* or web mail* or webmail* or Internet mail* or mailing list* or discussion list* or listserv*).tw.

5. ((patient or health or information or web or internet) adj portal*).tw.

6. (patient adj (web* or internet)).tw.

7. ((web* or internet or www or electronic* or online) adj5 (messag* or communicat* or transmi* or transfer* or send* or deliver* or feedback or letter* or interactiv* or input* or forum or appointment* or booking* or remind* or referral* or consult* or prescri*)).tw.

8. ((online or web* or internet) adj4 (service* or intervention* or therap* or treatment* or counsel*)).tw.

9. (e‐communication* or e‐consult* or e‐visit* or e‐referral* or e‐booking* or e‐prescri*).tw.

10. or/2‐9

11. physician patient relations/

12. professional patient relations/

13. interprofessional relations/

14. remote consultation/

15. or/11‐14

16. internet/

17. 15 and 16

18. 10 or 17

19. randomized controlled trial.pt.

20. controlled clinical trial.pt.

21. random*.tw.

22. placebo*.tw.

23. drug therapy.fs.

24. trial.tw.

25. groups.tw.

26. clinical trial.pt.

27. evaluation studies.pt.

28. research design/

29. follow up studies/

30. prospective studies/

31. (control* or prospectiv* or volunteer*).tw.

32. cross over studies/

33. comparative study.pt.

34. experiment*.tw.

35. time series.tw.

36. (pre test or pretest or post test or posttest).tw.

37. (pre intervention or preintervention or post intervention or postintervention).tw.

38. (impact* or intervention* or chang*).tw.

39. effect?.tw.

40. or/19‐39

41. exp animals/ not humans.sh.

42. 40 not 41

43. 18 and 42

Appendix 3. EMBASE (OvidSP) search strategy

1. e‐mail/

2. (electronic mail* or email* or e‐mail* or web mail* or webmail* or internet mail* or mailing list* or discussion list* or listserv*).tw.

3. ((patient or health or information or web or internet) adj portal*).tw.

4. (patient adj (web* or internet)).tw.

5. ((web* or internet or www or electronic* or online) adj5 (messag* or communicat* or transmi* or transfer* or send* or deliver* or feedback or letter* or interactiv* or input* or forum or appointment* or booking* or scheduling or remind* or referral* or consult* or prescri*)).tw.

6. ((online or web* or internet) adj4 (service* or intervention* or therap* or treatment* or counsel*)).tw.

7. (e‐communication* or e‐consult* or e‐visit* or e‐referral* or e‐booking* or e‐prescri*).tw.

8. or/1‐7

9. doctor patient relation/

10. interpersonal communication/

11. human relation/

12. patient counseling/

13. exp telemedicine/

14. telecommunication/

15. exp diagnostic test/

16. or/9‐15

17. internet/

18. 16 and 17

19. 8 or 18

20. randomized controlled trial/

21. single blind procedure/ or double blind procedure/

22. crossover procedure/

23. random*.tw.

24. trial.tw.

25. placebo*.tw.

26. ((singl* or doubl*) adj (blind* or mask*)).tw.

27. (experiment* or intervention*).tw.

28. (pre test or pretest or post test or posttest).tw.

29. (preintervention or postintervention).tw.

30. (cross over or crossover or factorial* or latin square).tw.

31. (assign* or allocat* or volunteer*).tw.

32. (control* or compar* or prospectiv*).tw.

33. (impact* or effect? or chang* or evaluat*).tw.

34. time series.tw.

35. or/20‐34

36. nonhuman/

37. 35 not 36

38. 19 and 37

Appendix 4. PsycINFO (OvidSP) search strategy

1. exp electronic communication/

2. (electronic mail* or email* or e‐mail* or web mail* or webmail* or internet mail* or mailing list* or discussion list* or listserv*).tw.

3. ((patient or health or information or web or internet) adj portal*).tw.

4. (patient adj (web* or internet)).tw.

5. ((web* or internet or www or electronic* or online) adj5 (messag* or communicat* or transmi* or transfer* or send* or deliver* or feedback or letter* or interactiv* or input* or forum or appointment* or booking* or schedul* or remind* or referral* or consult* or prescri*)).tw.

6. ((online or web* or internet) adj4 (service* or intervention* or therap* or treatment* or counsel*)).tw.

7. online therapy/

8. (e‐communication* or e‐consult* or e‐visit* or e‐referral* or e‐booking* or e‐prescri*).tw.

9. or/1‐8

10. exp therapeutic processes/

11. interpersonal communication/

12. telemedicine/

13. feedback/

14. or/10‐13

15. internet/

16. exp internet usage/

17. 15 or 16

18. 14 and 17

19. 9 or 18

20. ("32" or "33" or "34").cc.

21. (health* or medic* or patient* or clinic* or hospital* or illness* or disease* or disorder* or therap* or physician* or doctor* or psychotherap* or psychiatr* or telemedic* or treatment* or consult* or counsel* or referral* or remind* or appointment* or booking* or schedul* or visit* or prescri* or promot* or prevent* or diagnos* or test result* or screen* or intervention* or care).ti,ab,hw,id.

22. 20 or 21

23. 19 and 22

24. random*.ti,ab,hw,id.

25. (experiment* or intervention*).ti,ab,hw,id.

26. trial*.ti,ab,hw,id.

27. placebo*.ti,ab,hw,id.

28. groups.ab.

29. ((singl* or doubl* or trebl* or tripl*) and (blind* or mask*)).ti,ab,hw,id.

30. (pre test or pretest or post test or posttest).ti,ab,hw,id.

31. (preintervention or postintervention).ti,ab,hw,id.

32. (cross over or crossover or factorial* or latin square).ti,ab,hw,id.

33. (assign* or allocat* or volunteer*).ti,ab,hw,id.

34. (control* or compar* or prospectiv*).ti,ab,hw,id.

35. (impact* or effect? or chang* or evaluat*).ti,ab,hw,id.

36. time series.ti,ab,hw,id.

37. exp experimental design/

38. ("0430" or "0450" or "0451" or "1800" or "2000").md.

39. or/24‐38

40. limit 39 to human

41. 23 and 40

Appendix 5. ERIC (CSA) search strategy

(KW=(computer mediated communication* or electronic mail* or email* or e‐mail* or web mail* or webmail* or internet mail* or mailing list* or discussion list* or listserv*) or KW=((patient or health or information or web or internet) within 1 portal*) or KW=(patient within 1 (web* or internet)) or KW=((web* or internet or www or electronic* or online or on‐line) within 5 (messag* or communicat* or transmi* or transfer* or send* or deliver* or feedback or letter* or interactiv* or input* or forum or appointment* or booking* or schedul* or remind* or referral* or consult* or prescri*)) or KW=((online or on‐line or web* or internet) within 4 (service* or intervention* or therap* or treatment* or counsel*)) or KW=(e‐communication* or e‐consult* or e‐visit* or e‐referral* or e‐booking* or e‐prescri*)) and (KW=(health* or medic* or patient* or clinic* or hospital* or illness* or disease* or disorder* or therap* or physician* or doctor* or psychotherap* or psychiatr* or telemedic* or treatment* or consult* or counsel* or referral* or remind* or appointment* or booking* or schedul* or visit* or prescri* or promot* or prevent* or diagnos* or test result* or screen* or intervention* or care)) and (KW=(random* or trial* or placebo* or assign* or allocat* or volunteer* or crossover or cross over or factorial* or singl* blind* or doubl* blind* or clinical stud* or longitudinal stud* or control* or compar* or intervention* or preintervention or postintervention or pre test or pretest or post test or posttest or experiment* or prospectiv* or chang* or evaluat* or impact* or effect* or time series))

Appendix 6. CENTRAL (The Cochrane Library) search strategy

#1	MeSH descriptor Electronic Mail, this term only
#2	(electronic‐mail* or email* or e‐mail* or web‐mail* or webmail* or internet‐mail* or mailing‐list or discussion‐list or listserv*):ti,ab,kw
#3	(patient or health or information or web or internet) next portal
#4	patient next (web or internet)
#5	(web* or internet or www or electronic* or online or on‐line) near (messag* or communicat* or transmi* or transfer* or send* or deliver* or feedback or letter or interactiv* or input* or forum or appointment or booking or schedul* or remind* or referral or consult* or prescri*)
#6	(online or on‐line or web* or internet) near (service or intervention or therap* or treatment or counsel*)
#7	e‐communication or e‐consult* or e‐visit or e‐referral or e‐booking or e‐prescri*
#8	MeSH descriptor Computer Communication Networks, this term only
#9	(#8), from 1996 to 2002
#10	(#1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #9)
#11	MeSH descriptor Physician‐Patient Relations, this term only
#12	MeSH descriptor Professional‐Patient Relations, this term only
#13	MeSH descriptor Interprofessional Relations, this term only
#14	"doctor patient relation":kw
#15	"interpersonal communication":kw
#16	"human relation":kw
#17	"patient counseling":kw
#18	MeSH descriptor Telemedicine explode all trees
#19	telehealth or telemedicine or teleconsultation or telecommunication
#20	diagnostic‐test or laboratory‐test
#21	(#11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20)
#22	internet:kw,ti
#23	(#21 AND #22)
#24	(#10 OR #23)
#25	(#24)…………….[in Clinical Trials]

Appendix 7. CINAHL (EbscoHOST) search strategy

Search conducted by Cochrane Consumers and Communication Review Group and results sent to us.

Data and analyses

Comparison 1. Email versus standard mail: Primary outcome, Patient or caregiver behaviours/actions.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 During study, communicated with doctor by email (private access arm only)	1	52	Odds Ratio (M‐H, Fixed, 95% CI)	32.2 [1.78, 583.52]
2 Information requested from doctor by email about a disease or condition (private access arm only)	1	52	Odds Ratio (M‐H, Fixed, 95% CI)	14.16 [0.75, 265.85]
3 Information requested from doctor by email about appointments, test results or prescriptions (private access arm only)	1	52	Odds Ratio (M‐H, Fixed, 95% CI)	14.16 [0.75, 265.85]
4 Information requested from doctor by email about 'other' topic (private access arm only)	1	52	Odds Ratio (M‐H, Fixed, 95% CI)	0.85 [0.11, 6.51]
5 Discussed CRCS with no‐one (private access arm only)	1	52	Odds Ratio (M‐H, Fixed, 95% CI)	0.29 [0.05, 1.67]
6 Discussed CRCS with doctor (private access arm only)	1	52	Odds Ratio (M‐H, Fixed, 95% CI)	2.0 [0.49, 8.15]
7 Discussed CRCS with 'other' (private access arm only)	1	52	Odds Ratio (M‐H, Fixed, 95% CI)	0.29 [0.07, 1.29]
8 % Programme attendance	1	44	Mean Difference (IV, Fixed, 95% CI)	‐0.30 [‐10.95, 10.35]

Comparison 2. Email versus standard mail: Primary outcome, Patient or caregiver understanding and support.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Transformed scores for MOS Social Support: emotional/informational support	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
2 Transformed scores for MOS Social Support: tangible support	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
3 Transformed scores for MOS Social Support: Affectionate support	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
4 Transformed scores for MOS Social Support: Positive social interaction	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
5 Transformed scores for MOS Social Support: Overall support index	1	44	Mean Difference (IV, Fixed, 95% CI)	‐3.63 [‐12.67, 5.41]
6 Sallis Social Support and Exercise Survey Scores: participation: friends	1	44	Mean Difference (IV, Fixed, 95% CI)	3.59 [‐0.68, 7.86]
7 Sallis Social Support and Exercise Survey Scores: participation: family	1	44	Mean Difference (IV, Fixed, 95% CI)	2.27 [‐3.72, 8.26]
8 Sallis Social Support and Exercise Survey Scores: participation: others	1	44	Mean Difference (IV, Fixed, 95% CI)	0.08 [‐4.48, 4.64]
9 Sallis Social Support and Exercise Survey Scores: rewards and punishment: family	1	44	Mean Difference (IV, Fixed, 95% CI)	0.25 [‐0.65, 1.15]
10 Sallis Social Support and Exercise Survey Scores: rewards and punishment: others	1	44	Mean Difference (IV, Fixed, 95% CI)	‐0.59 [‐1.41, 0.23]
11 Likert‐type scale: participatory intention scores: 6 months	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
12 Likert‐type scale: participatory intention scores: 12 months	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

2.1. Analysis.

Comparison 2 Email versus standard mail: Primary outcome, Patient or caregiver understanding and support, Outcome 1 Transformed scores for MOS Social Support: emotional/informational support.

2.2. Analysis.

Comparison 2 Email versus standard mail: Primary outcome, Patient or caregiver understanding and support, Outcome 2 Transformed scores for MOS Social Support: tangible support.

2.3. Analysis.

Comparison 2 Email versus standard mail: Primary outcome, Patient or caregiver understanding and support, Outcome 3 Transformed scores for MOS Social Support: Affectionate support.

2.4. Analysis.

Comparison 2 Email versus standard mail: Primary outcome, Patient or caregiver understanding and support, Outcome 4 Transformed scores for MOS Social Support: Positive social interaction.

2.5. Analysis.

Comparison 2 Email versus standard mail: Primary outcome, Patient or caregiver understanding and support, Outcome 5 Transformed scores for MOS Social Support: Overall support index.

2.6. Analysis.

Comparison 2 Email versus standard mail: Primary outcome, Patient or caregiver understanding and support, Outcome 6 Sallis Social Support and Exercise Survey Scores: participation: friends.

2.7. Analysis.

Comparison 2 Email versus standard mail: Primary outcome, Patient or caregiver understanding and support, Outcome 7 Sallis Social Support and Exercise Survey Scores: participation: family.

2.8. Analysis.

Comparison 2 Email versus standard mail: Primary outcome, Patient or caregiver understanding and support, Outcome 8 Sallis Social Support and Exercise Survey Scores: participation: others.

2.9. Analysis.

Comparison 2 Email versus standard mail: Primary outcome, Patient or caregiver understanding and support, Outcome 9 Sallis Social Support and Exercise Survey Scores: rewards and punishment: family.

2.10. Analysis.

Comparison 2 Email versus standard mail: Primary outcome, Patient or caregiver understanding and support, Outcome 10 Sallis Social Support and Exercise Survey Scores: rewards and punishment: others.

Comparison 3. Email versus standard mail: Primary outcome, Health service, uptake of preventive screening.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 No. of participants having colorectal screening (completing either a FOBT, sigmoidoscopy or colonoscopy)	2	1012	Odds Ratio (M‐H, Fixed, 95% CI)	0.93 [0.69, 1.24]
2 No. of participants who had an mammography (Mayo Clinic employees)	1	847	Odds Ratio (M‐H, Fixed, 95% CI)	1.22 [0.91, 1.63]
3 No. of participants who had a Papanicalaou smear (Mayo Clinic employees)	1	847	Odds Ratio (M‐H, Fixed, 95% CI)	1.0 [0.74, 1.35]
4 No. of participants having colorectal screening (Mayo Clinic employees)	1	847	Odds Ratio (M‐H, Fixed, 95% CI)	1.25 [0.95, 1.65]
5 No. of participants having influenza vaccine (Mayo Clinic employees)	1	847	Odds Ratio (M‐H, Fixed, 95% CI)	1.25 [0.95, 1.66]
6 No. of participants having tetanus vaccine (Mayo Clinic employees)	1	847	Odds Ratio (M‐H, Fixed, 95% CI)	1.02 [0.72, 1.46]
7 No. of participants having lipid screening (Mayo Clinic employees)	1	847	Odds Ratio (M‐H, Fixed, 95% CI)	0.78 [0.53, 1.15]

3.2. Analysis.

Comparison 3 Email versus standard mail: Primary outcome, Health service, uptake of preventive screening, Outcome 2 No. of participants who had an mammography (Mayo Clinic employees).

3.3. Analysis.

Comparison 3 Email versus standard mail: Primary outcome, Health service, uptake of preventive screening, Outcome 3 No. of participants who had a Papanicalaou smear (Mayo Clinic employees).

3.4. Analysis.

Comparison 3 Email versus standard mail: Primary outcome, Health service, uptake of preventive screening, Outcome 4 No. of participants having colorectal screening (Mayo Clinic employees).

3.5. Analysis.

Comparison 3 Email versus standard mail: Primary outcome, Health service, uptake of preventive screening, Outcome 5 No. of participants having influenza vaccine (Mayo Clinic employees).

3.6. Analysis.

Comparison 3 Email versus standard mail: Primary outcome, Health service, uptake of preventive screening, Outcome 6 No. of participants having tetanus vaccine (Mayo Clinic employees).

3.7. Analysis.

Comparison 3 Email versus standard mail: Primary outcome, Health service, uptake of preventive screening, Outcome 7 No. of participants having lipid screening (Mayo Clinic employees).

Comparison 4. Email versus usual care: Primary outcome, patient or caregiver behaviours/actions.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 No. of visits to prescribed website within 1 week of clinic visit	1	83	Odds Ratio (M‐H, Fixed, 95% CI)	2.99 [1.17, 7.65]
2 Eating breakfast daily	1	49	Odds Ratio (M‐H, Fixed, 95% CI)	2.44 [0.53, 11.17]
3 Drinking alcohol in last month	1	49	Odds Ratio (M‐H, Fixed, 95% CI)	0.56 [0.18, 1.76]
4 Eating low fat food regularly	1	49	Odds Ratio (M‐H, Fixed, 95% CI)	0.5 [0.04, 5.91]
5 Eating low sugar items regularly	1	49	Odds Ratio (M‐H, Fixed, 95% CI)	0.87 [0.26, 2.95]
6 Median fruit and vegetable intake (portion number)			Other data	No numeric data
7 Median exercise episodes/week			Other data	No numeric data

4.3. Analysis.

Comparison 4 Email versus usual care: Primary outcome, patient or caregiver behaviours/actions, Outcome 3 Drinking alcohol in last month.

4.4. Analysis.

Comparison 4 Email versus usual care: Primary outcome, patient or caregiver behaviours/actions, Outcome 4 Eating low fat food regularly.

Comparison 5. Email versus usual care: Primary outcome, patient health status and well‐being.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Below weight maintenance band	1	49	Odds Ratio (M‐H, Fixed, 95% CI)	4.38 [0.45, 42.39]
2 In weight maintenance band	1	49	Odds Ratio (M‐H, Fixed, 95% CI)	1.5 [0.48, 4.65]
3 Above weight maintenance band	1	49	Odds Ratio (M‐H, Fixed, 95% CI)	0.37 [0.11, 1.26]
4 Mean % weight loss maintained at 6 months	1	49	Mean Difference (IV, Fixed, 95% CI)	2.80 [0.24, 5.36]
5 Median body weight (kg) maintained at 6 months (IQR)			Other data	No numeric data
6 Median amount of weight loss (kg) maintained at 6 months (IQR)			Other data	No numeric data

5.1. Analysis.

Comparison 5 Email versus usual care: Primary outcome, patient health status and well‐being, Outcome 1 Below weight maintenance band.

5.2. Analysis.

Comparison 5 Email versus usual care: Primary outcome, patient health status and well‐being, Outcome 2 In weight maintenance band.

5.3. Analysis.

Comparison 5 Email versus usual care: Primary outcome, patient health status and well‐being, Outcome 3 Above weight maintenance band.

Comparison 6. Email versus usual care: Primary outcome, patient or caregiver understanding and support.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Median perceived effort taken to control diet (IQR)			Other data	No numeric data
2 Median perceived effort taken to undertake activity (IQR)			Other data	No numeric data
3 Median perceived effort taken to maintain weight (IQR)			Other data	No numeric data

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Chan 2008.

Methods	Study design: Randomised controlled trial (pilot) Follow‐up: For the intervention arm follow‐up was approximately 2 months after final intervention contact and for the control arm it was approximately 2 months after enrolment. Recruitment: Research assistant approached patients in the waiting room and used a standardised script to ask if they were interested in participating in a research study about communicating with their doctor via email and if so administered an eligibility survey.
Participants	Description: Patients attending an outpatient General Internal Medicine Clinic. Setting: The outpatient General Internal Medicine Clinic, University of Texas‐Houston Clinic. Inclusions: 50 years or older, have at least a 6th grade level of education, have attended the outpatient General Internal Medicine Clinic at the university of Texas‐Houston clinic for at least a year, be due for colon cancer screening (CRCS) (i.e. no fecal occult blood testing within the past year, no flexible sigmoidoscopy, colonoscopy, or double‐contrast barium enema within the past 5 years), have a telephone, have private access to email and Internet or have an interest in access through the public library system and have their own transportation or be able to access public transportation Exclusions: Prior history of colorectal cancer or surgery Intervention group n = 42 from private access arm and n = 11 from the public access arm and control group n = 35 from private access arm and n = 9 from the public access arm.
Interventions	Intervention: Participants given a computer skills test, those passing test given log on details for the secure health network email system, those not passing scheduled for a tutorial (public library or home). A further tutorial was available if required. Two weeks after recruitment and/or the tutorial participants received NetLET (InterNET LETter) which consisted of a personalized email message from the participant’s primary care physician reminding them to undergo CRCS, informing them about their last CRCS test result, and providing a link to a secure web page with more information about CRCS. They were also mailed a fecal occult blood test (FOBT) kit via regular mail. If no response was received within a week the NetLET was emailed a second and third time at 1‐week intervals. Control: Two weeks after enrolment, the control groups were mailed a personalized letter signed by their physician with information about their last CRCS test, if relevant, a reminder about undergoing CRCS, and a FOBT kit. A second reminder letter was mailed about 2 weeks later without the test kit.
Outcomes	Self‐reported computer, Internet, and email use (via 2 month follow‐up survey, private access arm only). Information seeking behaviours (via 2 month follow‐up survey, private access arm only). Use of MyCareLink and NetLET (only for intervention groups and during intervention period). Number of participants returning the FOBT kit (during the intervention period).
Notes	All participants were given $20 upon completing the enrolment survey and were sent a $55 money order when they returned the follow‐up survey.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Participants were assigned to the private or public access arms depending on email and Internet access and then both arms were randomised into the intervention and control groups using random number assignment.
Allocation concealment (selection bias)	High risk	Separate sets of sealed envelopes were used to randomly assign participants to the intervention or control group however authors were unable to confirm whether these used appropriate safeguards.
Blinding (performance bias and detection bias) All outcomes	High risk	It is unclear whether outcome assessors were blinded however after seeking clarification from author, it is now known that the research assistant assigning the groups was aware of who was being assigned to the control and intervention groups. Due to the nature of the intervention participants could not be blinded and therefore knew if they were receiving the intervention or not. Similarly the physicians could not be blinded as they sent a personalised email message to patients, who were recommended to contact their physician to discuss test options.
Incomplete outcome data (attrition bias) All outcomes	Low risk	The incomplete outcome data were only adequately addressed for one outcome, which was the number of participants returning an FOBT kit ‐ for all those originally randomised to control and intervention in both arms (private and public) but not for the follow‐up survey assessing self‐reported computer, Internet, email use and information seeking behaviours. Overall response rate for the follow‐up survey was 61% and response rate was higher in the control group. Private access: I = 69%, C = 86% Public access: I = 9%, C = 67%. No investigation of non‐responders ‐ it was unclear why 39% of the participants did not respond.
Selective reporting (reporting bias)	High risk	Baseline data are presented for the private versus the public arm rather than for the control versus intervention arm. Authors state that they did compare intervention and control groups but these data are not presented. The authors did not analyse follow up data on the public access participants and state that this is because of the low response rate to the survey in the intervention group.
Other bias	High risk	Baseline comparability: study text states that the only significant difference between groups at baseline was for education in the public access arm where the control group participants were more likely to have had a less formal education. Validity of measure: No information given on whether baseline and follow‐up questionnaires are validated. Recall bias: self‐reported questionnaires could over or underestimate results Selection bias: within the public access arm only those who were interested in accessing the Internet and email through the public library system were recruited.

Chaudhry 2007.

Methods	Study design: Randomised controlled trial Study duration: 12 months Recruitment: Patients due for annual mammography in the next three months were identified from the Primary Care Physician Portal (web‐based information sytem at Mayo Clinic Rochester).
Participants	Setting: The Division of Primary Care Internal Medicine (PCIM), Rochester , Minnesota, U.S. Inclusions: Female patients of the PCIM between the ages of 40 and 75 years, all due for an annual mammography in the next three months. Exclusions: Having had mammograms performed within the system in the prior nine months. Intervention: 3326, Control: 3339. Subgroup: Email: 399, Mail: 448.
Interventions	Initially randomised into intervention and control groups. Intervention: three months before they were due for annual screening, participants received a personalised letter via US mail from their physician indicating the need for screening mammography and advising them to call to schedule an appointment with a brochure that explained all adult preventive services. Control: received usual care. Then a subset of females who were Mayo Clinic employees were randomly assigned to either mail or email groups. Both then received a reminder for screening on behalf of their primary provider, the email group received this via their work email account.
Outcomes	Mammography screening rates (at end of 12 month intervention period). Other preventive screening rates (Papanicolaou smear, colorectal screening, pneumonia vaccine, influenza vaccine, tetanus vaccine and lipid screening) (at end of 12 month intervention period). Use of annual physical examinations in preceding 12 months (determined from the administrative data at end of 12 month intervention period).
Notes	For the purposes of this review we are interested in the second level of this study where email reminders are compared to mail reminders in a subset of Mayo Clinic employees.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Participants were randomly assigned using computer generated random numbers.
Allocation concealment (selection bias)	Unclear risk	The method of allocation concealment remained unclear even after author contact.
Blinding (performance bias and detection bias) All outcomes	High risk	Additional information provided by the author indicated that physicians were unaware of group assignment. However the appointment secretaries responsible for sending the email and letters were given participant names and therefore were not blinded. Although not clearly stated it is presumed that due to the nature of the intervention participants could not be blinded.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	It is unclear whether an intention to treat analysis was carried out, the report states that some women declined to consent after randomisation and it is unclear what happened to these participants in the analysis.
Selective reporting (reporting bias)	High risk	An additional logistic regression model was carried out (for the Mayo Clinic employees group only) but this was not pre‐specified in the methods section of the report.
Other bias	High risk	Baseline comparability: no baseline data presented for any participants. Validity of measure: Data collected from administrative systems and therefore it may be difficult to validate routinely collected data. Selection bias: only Mayo Clinic employees were randomised to receive email or letter reminders and not the whole sample.

Muller 2010.

Methods	Study design: Randomised controlled trial. Study duration: 90 days. Recruitment: The Health Maintenance Organisation's research centre identified 18,847 patients who were both enrolled in the secure email service and were due for colorectal cancer (CRC) screening.
Participants	Setting: nonprofit Health Maintenance Organisation (North West USA) with approximately 479,000 members. Inclusions: men and women aged between 50 and 80 who had accounts on the secure email system, had 12 months or more insurance coverage with no more than one 45 day break prior to November 1 2007 and due for colorectal screening. Exclusions: patients who had fecal occult blood testing (FOBT) in the previous 12 months, a sigmoidoscopy in the previous 5 years or a colonoscopy in the previous 10 years as recorded in their medical record, patients who had a total colectomy, a history of colon cancer, or inflammatory disease, use of anticoagulants or an oncology visit in the previous 12 months but had not yet had the procedure, patients for whom a FOBT had been ordered in the previous 3 months, patients who were on hospice or in a nursing home facility and those with a diagnosis of dementia. 2100 potentially eligible participants identified, 691 found to be ineligible on second screen. Remainder randomised into: email: 457, letter: 458, control: 494.
Interventions	Two intervention arms. Letter: received a single letter reminder. Email: received a single email reminder delivered through the secure email system inviting patients to pick up an FOBT kit at the lab for colorectal screening to test at home. Content of the letter and email were identical. Control: received usual care.
Outcomes	Subjects receiving colorectal screening (ninety days after reminder sent by reviewing participant charts). Subjects not receiving colorectal screening (ninety days after reminder sent by reviewing participant charts).
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Participants were randomised into one of three arms and were matched for age and sex but no further information is given and we were unable to contact the authors.
Allocation concealment (selection bias)	Unclear risk	Insufficient information provided and we were unable to contact the authors.
Blinding (performance bias and detection bias) All outcomes	High risk	The researcher responsible for sending and chasing up reminders could not be blinded to group allocation. Participants were not informed that they were part of the study and did not consent, therefore they were 'blind' to being in the study rather than to which group they were in.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All incomplete data have been adequately addressed. An intention to treat analysis was carried out.
Selective reporting (reporting bias)	Low risk	No evidence of selective reporting, the results reported match those stated in the methods section.
Other bias	Unclear risk	Baseline comparability: age and gender were compared and found to be comparable Validity of measure: An information system is used to obtain information about participants and it is unclear whether validation is relevant. Measurement bias: no information given on how positive response rates were recorded.

Pearson 2004.

Methods	Study design: Randomised controlled trial. Study duration: 6 weeks. Recruitment: sample of participants drawn from a previously identified maintenance cardiac rehabilitation programme.
Participants	Setting: maintenance cardiac rehabilitation programme based in hospital‐owned health and fitness facility, Florence, Northeastern South Carolina. Inclusions: cardiac rehabilitation participants self‐identified as computer owners/users. No exclusion criteria. 75 patients eligible, 52 participants who self identified as online computer owners/users were put into the intervention group and randomised into email and standard mail groups, and the remaining 23 into the control (computer users but not online).
Interventions	Two intervention arms. Email: received fitness tips delivered electronically via email. Standard mail: received the fitness tips via standard mail. The content of the fitness tips was consistent across the two intervention groups and both groups received one mailing per week for a total of 6 fitness tips for 6 weeks duration. An attempt was made for both groups to receive to receive the same information simultaneously. Control: received usual care and were given the fitness tips at the end of the study period.
Outcomes	Levels of perceived availability of functional support (via survey in the week following the intervention period) Perceived availability of social support for exercise; (via survey in the week following the intervention period) Participatory intentions of study participants (via survey in the week following the intervention period) Cardiac rehabilitation programme attendance (at end of intervention period using programme check in data and verified by participant logs).
Notes	Prior to this study, an initial pilot study was conducted first among participants in a hospital based maintenance cardiac rehabilitation programme with the same participants used in this study. This study is described by the study author as quasi‐experimental although it has a mixed design. Participants self‐identified as 'online' computer users were designated for the intervention arm of the study and were randomly assigned to either email or standard mail group. This is the part of the study that qualifies as a RCT and the email and mail arms are the only groups of interest for the purpose of this review. We will not be using the data from the control arm as this was not randomised.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	The initial split of 'intervention' and 'control' groups was not randomised. However the 'intervention' group were then randomised into email and mail groups using a random number table although insufficient information is provided on the sequence generation.
Allocation concealment (selection bias)	High risk	This was not clear in the methodology and author confirmed on contact that the allocation sequence was not concealed.
Blinding (performance bias and detection bias) All outcomes	High risk	Author confirmed that both investigators and participants were not blinded to their group allocation.
Incomplete outcome data (attrition bias) All outcomes	High risk	Numbers and reasons are given for drop‐outs but these are not specified for the email and standard mail groups of the intervention arm. An intention to treat analysis was not carried out.
Selective reporting (reporting bias)	High risk	Some tests for significant differences are reported as non‐significant but no P values are reported.
Other bias	High risk	Baseline comparability: Data are presented comparing the email, standard and control groups but the two randomised groups (email and mail) were not directly compared and so we cannot tell if the two groups were comparable. Validity of measure: Medical Outcomes Study Social Support Survey and the Sallis Social Support and Exercise Survey were validated and the participatory intentions was based on the construct of behavioral intentions found in the theory of reasoned action and the theory of planned behaviour. Selection bias: only selecting cardiac rehabilitation participants who are 'online' computer users for the intervention group therefore limiting generalisability. Reporting bias: participants may have felt a need to please the investigator after developing an ongoing relationship and being aware they were part of a trial. Recall bias: self‐administered questionnaires may introduce bias.

Ritterband 2005.

Methods	Study design: Randomised controlled trial. Study duration: 1 week. Recruitment: Families were approached if they had a child who was being seen for the first time in the Pediatric Gastroenterology clinic at the University of Virginia with a chief complaint of chronic constipation and/or encopresis.
Participants	Setting: Pediatric Gastroenterology clinic at the University of Virginia. Inclusions: Families (caregivers) of children being seen for the first time at the Pediatric Gastroenterology clinic with a chief complaint of chronic constipation and/or encopresis. Families had to have access to the Internet in their home and have an active email account. No exclusion criteria. Participants randomised into: intervention: 43, control: 40.
Interventions	During clinic visit, physicians instructed all families to visit a website that provided educational information relevant to their child's problem. Families were given a form with the Web‐site address and a log‐in identification number. Intervention: Two days after their clinic visit, half of the families received an email reminding them to visit the web site. Control: were not sent an email reminder to visit the web site (''no prompt'' group).
Outcomes	Number of visits to prescribed website within one week of clinic visit with or without email reminders (website automatically logged this information). Experiences accessing website and identification of barriers to access (reported for a proportion of the whole sample and not by intervention/control group) (via telephone or email from researcher 1 week after clinic visit).
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Assigned randomly using a chart, however we have no further information even after author contact.
Allocation concealment (selection bias)	Unclear risk	Insufficient information provided even after author contact.
Blinding (performance bias and detection bias) All outcomes	Unclear risk	It is unclear whether the researcher was blinded. Physicians remained blinded to the study condition. It is unclear whether participants were aware if they were to receive a prompt.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Incomplete outcome data were adequately addressed for the number of visits to prescribed website but not for the experiences accessing the website (only 81% of participants were contacted and no further information is given on those who were unreachable). However, this outcome was not assessed as part of this review as the data were not disaggregated by the intervention and control arms. An intention to treat analysis was carried out for the number of visits to prescribed website.
Selective reporting (reporting bias)	High risk	The data for experiences accessing website and identification of barriers to access was presented for the whole group and not by intervention/control groups.
Other bias	High risk	Baseline comparability: no significant differences between the two groups on type and speed of Internet connection, number of times they reported checking their email, frequency of using the Internet and child's age. Validity of measure: no information given on whether use of the database/log in system was validated. No information on whether the follow‐up interview questionnaire. Selection bias: participants had to have access to the Internet at home. Responder bias: the participants may have answered differently over the telephone when asked by a researcher about their experiences. Recall bias: participants not using the website/using it less frequently may have been less able to recall their experiences.

Thomas 2008.

Methods	Study design: Randomised controlled trial. Study duration: 6 months. Recruitment: patients recruited directly from the twice weekly weight loss clinic.
Participants	Setting: weight loss clinic within Portsmouth Hospital Trust. Initially, patients attending the clinic had a Body Mass Index ≥ 30 kg/m2, with or without co‐morbidities. Inclusions: patients who had lost weight to a level of > 5% of initial body weight through attendance at a dietician‐led NHS outpatient clinic and about to be discharged from the clinic. Exclusions: Patients applying for bariatric surgery, without email access, or taking weight loss medication, Individuals with binge eating disorder and learning difficulties (they attend specialist dietetic clinics). Participants randomised into intervention: 28, and control groups; 27.
Interventions	Intervention: participants received 6 months of weekly email from the dietitian called ‘Tip of the Week’. Tips gave dietary, behavioural and exercise advice. Participants were sent a monthly email requesting a report of their current weight and asked whether they were managing to keep within their personal weight band which was 2kg either side of their weight at the start of the intervention. Control: received no contact from the dietitian but were asked to return to the clinic at the end of the study period.
Outcomes	Maintenance of weight loss within a defined range (face‐to‐face meeting with researcher at end of intervention period, weight recorded) Perceived effort in achieving weight maintenance (self‐administered questionnaire during researcher meeting at end of intervention period). Specific dietary changes (fruit and vegetable intake, alcohol intake, eating low fat and low sugar foods, eating breakfast) (self‐administered questionnaire during researcher meeting at end of intervention period). Ability to maintain a level of activity (exercise episodes per week) (self‐administered questionnaire during researcher meeting at end of intervention period). Cost effectiveness (price per patient, intervention group only) (unclear how this was measured).
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer random number generator used to create sequence and allocation was made in groups of three to email and control groups alternately.
Allocation concealment (selection bias)	Low risk	Consecutively numbered sealed opaque envelopes were used.
Blinding (performance bias and detection bias) All outcomes	High risk	Researcher was not blinded: 'it was not possible to protect the trial against a bias by keeping the researcher unaware of the identity of the experimental group.' Participants were not blinded: those in the intervention group had to sign a consent form to confirm that they have been made aware of the insecure nature of electronic mail and therefore aware that they were receiving the intervention.
Incomplete outcome data (attrition bias) All outcomes	High risk	An intention to treat analysis was not carried out. No analysis was carried out on non‐completers in the control group with respect to their reasons for not completing, however the attrition rate was low (8%).
Selective reporting (reporting bias)	Low risk	There was no published protocol, but results are presented for each outcome as outlined in the methods section.
Other bias	High risk	Baseline comparability: 'The median body weight, BMI and percentage weight loss achieved by participants at entry to the study did not differ significantly between the two groups'. Validity of measure: Both questionnaires were formally validated (Likert scale and C96 questionnaire). Performance bias: the authors acknowledged some performance bias: six months may not have long enough to measure weight maintenance as stated 'ideally weight loss maintenance should take place for longer than 6 months'. Selection bias: Some of the included participants were not in maintenance phase and continued to lose weight (10%). They lost significant amounts of weight (>20% initial body weight). This may have biased results as these participants would have continued to lose weight regardless of the intervention. Reporting bias: the relationship with the dietician may have influenced participant reporting. Weight is a sensitive subject and participants may have felt the need to 'please' the dietician. Recall bias: This could have occurred during the exit interview which could have over or underestimated the results.

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Burgard 2006	Intervention not administered in a healthcare setting; participants recruited via university campus email system.
Liguori 2007	Intervention not administered in a healthcare setting; participants recruited at a conference.

Characteristics of ongoing studies [ordered by study ID]

Catz NCT00923624.

Trial name or title	Patient Portal to Support Treatment Adherence
Methods	RCT (parallel)
Participants	Inclusion Criteria: Male or female, aged 18 or older enrolment in GHC or KP health plans for at least 6 months expect to stay enrolled for 1 year in health plan HIV+ currently prescribed antiretroviral medication regular access to a computer with Internet capabilities and working email address and meet one of the following: gap in antiretroviral prescription refills based on 6 month refill records referral to study by provider team for adherence problems Exclusion Criteria: dementia or psychosis diagnosis documented in medical record
Interventions	300 adult HIV+ patients will be randomised to one of two arms. 150 will be assigned to the experimental arm: receiving secure messages focused on antiretroviral adherence from a nurse through the health plan EMR patient website. 150 will be assigned to the attention control comparison arm, and will receive electronic messages from a study staff member that provide information about the various features of the health plan patient website. No usual care arm.
Outcomes	Primary Outcome: Lapses in medication adherence over the last 7 and 30 days (self report) and over previous 12 months (pharmacy records). Secondary Outcome: Web service utilization (assessed at 12 months).
Starting date	June 2009
Contact information	Sheryl L. Catz, PhD. Group Health Co‐operative: catz.s@ghc.org
Notes	Currently recruiting participants. Estimated study completion date August 2012.

Morgan ACTRN1260900092524.

Trial name or title	Promotion of self‐help strategies for sub‐threshold depression: An e‐mental health randomised controlled trial
Methods	RCT (parallel)
Participants	Inclusion criteria: male or female, over 18 years of age resident of Australia, New Zealand, UK, Ireland, Canada or the USA access to the Internet at least weekly sub‐threshold symptoms of depression (two to four symptoms of depression, experienced for two or more weeks, that have affected work, home, or social functioning) Exclusion criteria: currently receiving treatment for depression from a professional
Interventions	Participants will receive 12 emails over 6 weeks. Each email will contain a self‐help strategy for coping with depressive symptoms. The email will contain information about why the strategy will be effective, tips for implementing the strategy and overcoming barriers, and how to set a goal to implement the strategy. Strategies are based on previous research published by the trial co‐ordinators.
Outcomes	Primary outcomes change in depressive symptoms as assessed using the Patient Health Questionnaire (PHQ‐9, self‐rated reduction in development of major depression, as assessed using the PHQ‐9 (self‐rated). Secondary outcomes: change in coping strategies used (researcher devised questionnaire) change in functioning as assessed using the Work and Social Adjustment Scale (self‐rated) change in knowledge of depression (researcher devised questionnaire) change in psychological distress as assessed using the K10 (self‐rated)
Starting date	1 January 2010
Contact information	Amy Morgan, Orygen Youth Health Research Centre Centre for Youth Mental Health, University of Melbourne, Locked Bag 10 Parkville, VIC, 3052, Australia, email: ajmorgan@unimelb.edu.au.
Notes	This study is not yet recruiting

Samuelsson NCT01032265.

Trial name or title	Web‐based Management of Female Stress Urinary Incontinence. Evaluation of a Treatment Programme With Pelvic Floor Muscle Training (PFMT) and Elements of Cognitive Behavioural Therapy
Methods	RCT ‐The aim of this study is to determine if web‐based management of female SUI, with a treatment using PFMT and elements of CBT is effective compared to treatment supported by a pamphlet. The duration of the treatment programme is three months, follow‐up at four months, 1 year and two years.
Participants	Inclusion criteria: female aged between 18 to 70 years stress urinary incontinence leakage once a week or more often ability to read and write Swedish asset to computer with Internet connection Exclusion criteria: pregnancy former incontinence surgery known malignancy in lower abdomen difficulties with passing urine visual blood in urine intermenstrual bleeding severe psychiatric diagnosis neurological disease with affection on sensibility in legs or lower abdomen
Interventions	Intervention arm to be given a web‐based treatment programme with PFMT and elements of cognitive behavioural therapy which includes regular email contact with urotherapist. Comparator arm to be given a pamphlet treatment which includes lifestyle information and PFMT exercises.
Outcomes	Primary Outcome Measures: Symptom scoring instrument (ICIQ‐UI Short Form) ( Time Frame: 4 months ) Health related Quality of Life (HrQoL), condition specific (ICIQ‐LUTSqol) ( Time Frame: 4 months ) Secondary Outcome Measures: Comparison of health economics ( Time Frame: 4 months ) Usage of incontinence aids ( Time Frame: 4 months ) Patient satisfaction ( Time Frame: 4 months ) Incontinence Episode Frequency (IEF) ( Time Frame: 4 months ) Patient`s Global Impression of improvement scale (PGI‐I) ( Time Frame: 4 months )
Starting date	December 2009
Contact information	Eva Samuelsson, MD, PhD, Umea University, Umea, Sweden, Email: eva.samuelsson@jll.se.
Notes	This study is currently recruiting participants.

Scrol NCT01077388.

Trial name or title	Collaborative Behavioral e‐Care to Decrease Cardiovascular Risk (e‐Compare)
Methods	Randomised trial
Participants	Inclusion criteria: all adults age 30‐<70 men and women who have received care at GH primary care medical centres in Western Washington and who have been continuously enrolled in GH for at least 2 years. BMI 26 or higher, uncontrolled HTN (based on the 2 most recent visits), and moderate risk for CVD. Exclusion criteria: a history of any type of cardiovascular disease (CVD). Including TIA, stroke, coronary heart disease (CHD), angina, congestive heart failure (CHF), peripheral arterial disease (PAD), ar, CHF, PAD), arrhythmias, cardiomyopathy, and valvular disease. pre existing diagnosis of diabetes diagnosis of a serious life threatening or limiting condition pre existing diagnosis of alcohol or chemical dependency pregnant or planning on becoming pregnant in the next 6 months. Women randomised to the intervention group who become pregnant during the study will be given the option of continuing the intervention in conjunction with usual pre‐natal care provided by GH, but will not be included in the final analysis for the study. Women randomised to the self care group who become pregnant during the study will be given the home monitoring equipment usually given to the control group at the end of the study. They will then be referred for usual care and will also not be included in the final analysis for the study. unable or unwilling to participate
Interventions	The intervention arm will receive a dietitian‐delivered behavioral intervention, that uses a patient shared EMR and e‐communications, and will be integrated into routine healthcare. The comparator arm will be self‐care where participants will continue to get care as usual from their regular doctor
Outcomes	Not given
Starting date	May 2010
Contact information	Aaron Scrol, MA, Group Health Cooperative, Seattle, Washington, United States, email: scrol.a@ghc.org
Notes	This study is not yet open for participant recruitment.

Differences between protocol and review

Types of participants, types of interventions, types of outcome measures and data extraction and management

The wording in these four sections has been changed to improve the clarity of who the senders and recipients of the email communications were, as well as to reflect that the communication is one‐way. The primary patient outcomes were also clarified in respect to caregiver outcomes.

Search methods for identification of studies

It was stated in the protocol (Atherton 2009b) that we would search the following databases as part of the grey literature search:

• Dissertation Abstracts (North American and European theses) via British Library

• TrialsCentralTM (www.trialscentral.org)

We did not search these databases, and this decision was made in conjunction with the Cochrane Consumers and Communication Review Group. TrialsCentral TM was unsearchable as the website seemed only to pull information in from other sources. The only search options were to search by condition or intervention for clinical and drug interventions only (no free text). Dissertation Abstracts were not searched as several of the other databases would duplicate this search (Index to Theses and ProQuest).

MEDLINE search

Minor changes have been made to the MEDLINE strategy since the protocol stage. The new version of the search can be found in Appendix 2 of the review. These changes were made in conjunction with John Kis‐Rigo, Trials Search Coordinator at the Cochrane Consumers and Communication Review Group. The changes involve the removal of the term 'on‐line' from the strategy. This is because MEDLINE changed the way it processed this term, and we were retrieving a very high number of articles (20,000+) where as before the change in processing we had obtained around 8000. Removing this term brought the retrieval rate back to acceptable levels.

Assessment of heterogeneity

This section has been amended to reflect most recent guidance from the Cochrane Collaboration on assessing heterogeneity, and to bring it in line with the other four reviews on email interventions.

Data synthesis

This section has been amended to reflect the type of analysis possible, and to bring it in line with the other four reviews on email interventions.

Contributions of authors

Prescilla Sawmynaden assisted with the search, carried out data extraction and analysis and wrote the review.

Helen Atherton wrote the protocol, carried out the search, was data extractor, assisted in analysis and co‐wrote the review.

Azeem Majeed commented on drafts of the review.

Josip Car conceived the idea for the review and supervised the production.

Sources of support

Internal sources

• eHealth Unit, Department of Primary Care and Public Health, Imperial College, UK.

  The review received a partial financial contribution from The Department of Primary Care and Public Health, Imperial College London. The Department of Primary Care & Public Health at Imperial College is grateful for support from the NIHR Collaboration for Leadership in Applied Health Research & Care (CLAHRC) Scheme, the NIHR Biomedical Research Centre scheme, and the Imperial Centre for Patient Safety and Service Quality.

• Department of Family Medicine, University of Ljubljana, Slovenia.

  JC is a visiting researcher in the Department, receiving salary and office space support.

• NHS Education for Scotland, UK.

  BM was funded during the production of the protocols by NHS Education for Scotland.

• NHS Connecting for Health Evaluation Programme (NHS CFHEP 001), Not specified.

  http://www.haps.bham.ac.uk/publichealth/cfhep/

External sources

• Medical Research Council, UK.

  HA is the recipient of a Medical Research Council PhD Studentship, administered by Imperial College, London, UK.

Declarations of interest

None known.

New