Thirty to 40% of family practice patients have a mental health condition and family physicians are frequently their only contact, but easy access to satisfactory psychiatric backup is often difficult to obtain (1, 2). If a colleague is unavailable (3) most (64%) family practioners’ questions are not pursued (4) because they are complex and difficult to resolve with single text-based sources (5). This issue is infrequently solved by consultation, because consultants’ letters are often inadequate to the needs of family practitioners (6, 7).

Within mental health a variety of shared care initiatives have addressed this gap, with frequently positive results where attention was paid to communication, mutual education, and respect between the disciplines (8–10). However, covert barriers to integration remain, including financial pressures, the lack of shared models, and geographic separation (11). Therefore, any process that allows faster, more flexible, and more responsive clinically relevant communication would be beneficial.

The mentoring relationship could serve just such a function. In fact, the “learning relationship” characteristic of mentoring has many of the attributes of the ideal technique for supporting family practitioners (12). The mentoring relationship is “teaching taken to a deeper level,” which addresses the learning needs of the mentee in an attuned and respectful manner over time (13). This article reports on such a novel mentoring program, one aspect of which is the use of e-mail as a mode of communication between mentors and mentees.

The Collaborative Mental Health Care Network

Ontario has more than 12 million people living in an area of 415,000 square miles (14). Many family practitioners are located far from mental health services. The Collaborative Mental Health Care Network was founded in 2001 to link family practitioners with mental health resources by building teams of 10 to 25 family practitioners with a pair of mentors consisting of a psychiatrist and a general practice psychotherapist (a family practitioner whose practice centers on mental health care) (15). Functionally, this allows a family practitioner to bypass long waiting lists and access timely advice about patients. The Network has expanded to approximately 400 family practitioners and 66 mentors in regular communication by e-mail, telephone, small-group meetings, and an annual continuing medical education (CME) event. The provincial government provides funding, which is used for administration, to support the CME event, and to pay mentors. Liability concerns were addressed with the appropriate professional organizations, and they deemed the Network practice reasonable, with liability equivalent to a hallway conversation between colleagues.

With respect to the perceived usefulness of the Network, a previous survey indicated that high utilizers reported increases in collegiality and confidence in their ability to handle complex problems. They were also used in their communities as mentors to other family practitioners, implying a “trickle down” benefit to the community at large.

Given these data, we attempted to deepen our understanding by examining the communication of mentor/mentee dyads. We focused on e-mail because its use in the Network increased from 20% to over 40% from 2004 to 2006, suggesting an advantage to e-mail. E-mail also provides an accessible record of mentor-mentee communication, permitting the analysis of a novel medium. Our goal was to generate hypotheses to inform future research and Network development.

Two sources were utilized to describe Network functioning. The first, a new quality assurance survey, was administered in 2004. Overall, 137 mentees responded, representing 53% of the program at that time.

The second source consisted of a qualitative analysis of a convenience sample of e-mail correspondence. After approval by the Research Ethics Review Board of the principal investigator’s (JH) hospital, participants were invited to provide informed consent and forward samples of Network e-mail correspondence to the researchers. E-mails were received from eight mentors and 26 mentees, constituting 16% of those who report e-mail as a primary mode of communication. Because the sample is neither random nor representative, it does not allow for generalizability, but it provides rich material for generating hypotheses to be tested further in future focus groups or surveys.

The qualitative analysis examined all e-mails obtained (after removal of identifiers) to detect the themes that emerged and recurred. The analysis was guided by qualitative questions such as “What are the issues that most occupy the mentors and mentees? What are the functions of the Network? What does the Network do best? What kinds of relationships develop or appear between the mentors and mentees?” This technique, considered sensitive to emergent qualities (12, 13), is suited to this novel setting, where the mode of communication (e-mail) is new, precluding an a priori categorization of relevant features.

Results show that the primary functions of the Network are to provide clinical support and CME. There are both content and process aspects to this education. Content aspects address topics that the mentees identify as important, and process aspects address the means of acquiring that information and, more subtly, how the mentee/mentor relationship develops. A complete sample communication is included in Appendix 1 , and quotes from this are included in parentheses to illustrate these points. Lastly, there are some data on the effect of Network participation.

Content

Information regarding content was derived primarily from the survey. Mentors reported that the largest number of requests were for help with general, nonemergency clinical issues (70%, n=96), with topics relating to pharmacotherapy (53%, n=73), psychotherapy (34%, n=46), treatment review (27%, n=37), and diagnosis (24%, n=33). E-mail analysis suggested content teaching also occurs on the “how to” of practice, such as how to divide up the time of a psychotherapy session.

Process

E-mail analysis showed that knowledge transfer occurs both directly and indirectly. Direct knowledge transfer makes the information conveyed explicit (e.g., mentor: “start at 25 mg b.i.d. and increase to 50 mg b.i.d. after 1 week”). Indirect teaching occurs when the mentor encourages attaining knowledge via another means, such as studies and literature reviews or the involvement of another member of the Network whose experience may add further depth.

Mentors also typically encourage mentees to generalize knowledge. This occurs via modeling behavior, such as asking important questions that allow the mentee to see beyond the immediate problem. Additionally, mentors will often use language such as “I usually …” which encourages the mentee to apply these approaches beyond the specific case. For example, our mentor makes a point about nonadherence as a recurrent issue among patients with bipolar affective disorder. Even though it is not the most pressing issue with this individual, the information will contribute to the mentee’s management of a future patient with bipolar affective disorder.

Effective education occurs in the context of a trusted relationship that permits displays of ignorance or need without the risk of shame. Therefore, we wished to investigate how the mentee/mentor relationship developed. Both address aspects of the developing relationship via direct comments and attitudinal statements. For instance, mentees demonstrate an active need for help (e.g., mentee: “Please help”) and are grateful for advice. They also manifest a position that suggests a degree of vulnerability. This becomes clear with the use of apologies for lengthy histories, a deferential attitude toward mentors, frequent expressions of their own inability to care for the patient and substantial concern for the patient. In turn, mentors encourage the mentee. They are very reassuring about mentees’ choices of cases as legitimately difficult (e.g., mentor: “The diagnosis is difficult … This is a good start”).

In particular, mentees value mentors for their credibility as experts. The expertise is conveyed by the mentors explicitly invoking authority by referring to their experience. This expertise goes beyond content knowledge into attitudes or styles of practice. For example, the mentor in our example introduces the Ulysses maneuver, even though the mentee does not ask for it, and thus the mentor reveals his or her own timing of psychotherapeutic interventions. This expertise is crucial to the mentees’ valuing of the mentors, as shown by the ways in which mentees strive to keep the “hierarchy” in place. This occurs via self-deprecating humor, apologies for bothering the mentors, and of course, explicitly asking for advice. When faced with this “elevating” discourse, mentors respond by emphasizing equality in the communication. They are less formal than mentees and make frequent use of humbling conversational devices that appear to be a deliberate face-saving strategy to mitigate the apparent knowledge advantage.

Effect

That this hierarchy serves the purpose of effective education is revealed by the gratitude expressed by mentees (e.g., mentee: “We are very happy”). Survey results indicated satisfaction with the Network was high, with 88% of family practitioners reporting that their ability to provide mental health care had improved (n=120) and 75% that their knowledge base had increased significantly (n=103).

The Network effectively conveys clinical support and continuing medical education, with the participants reporting high satisfaction. The most effective medical educational strategies are clinically relevant, occur via multiple interventions over time, and are immediately applicable to practice (16). We believe that the Network meets these criteria in that the interactive, conversational style of case discussion conveys contextualized education that a family practitioner can readily translate into generalized knowledge.

The qualitative analysis also suggests that a trusted relationship with a colleague who has a superior and relevant knowledge base is critical to the Network’s effectiveness. This disparity in expertise results in a “knowledge hierarchy,” and in fact, many of the perceived consultative benefits derived from mentoring are a function of this hierarchy (17). Mentees value mentors who can provide experience and objectivity (18) but who are also “trustworthy, nonjudgmental, accessible, and reliable” (19). Thus, the knowledge hierarchy does not inevitably produce a power hierarchy. The mentoring relationship appears to develop through the direct expression of need and gratitude on the part of the mentee and the knowledge transfer and direct encouragement on the part of the mentor. The mentees appreciate and emphasize the mentors’ expertise, but the mentors counter this with face-saving, equalizing comments such as displaying their own need to look elsewhere for information. The result is a safe and secure interpersonal space in which educational needs or difficult patient encounters can be revealed, for the purpose of gaining insight, skills, or knowledge, without the risk of shame.

Medical education via e-mail has not been reported on in great detail (20–22). The benefits of e-mail compared with the standard consultation letter include the rapid response and the reduction of formality, both of which support the relational aspects of mentoring. Telephone calls can be burdensome because they must be accepted when they occur, even if in the course of a busy clinic, or booked formally. However, e-mail can be accessed as it suits the participants, perhaps after reflection or checking of references. E-mail also provides an immediate hard copy of the consultation for the chart, as opposed to the onerous extra step of transcription.

The drawbacks of e-mail include concerns about privacy and security, although sensitive information is easily removed. E-mail does not allow the same level of interactivity that a real-time conversation does, and the lack of prosody and paralinguistic features can make e-mail unnecessarily cumbersome. A myriad of possible technological glitches also may occur. Our sense is that both mentors and mentees develop an appreciation about how to communicate about what, and shift the mode of communication when necessary.

With respect to the limitations of this report, it is clearly neither a representative nor random sample, and our conclusions must be limited in scope. However, our goal was to qualitatively explore the way Network communication functions and to generate hypotheses for future validation. As such, it has directed our attention to contextualized education, the relevance of personalized dyadic or group relationships for learning, and the distinction between knowledge and power hierarchies. Lastly, although subjective reporting on the Network is positive, the lack of available data on objective change in the performance of family practitioners or patient outcomes makes it impossible to state the extent of change in actual patient care (23).

Grant support for this study occurred via the Ontario Ministry of Health and Long Term Care, which provides funding for the Collaborative Mental Health Care Network.