This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Medical Education, is properly cited. The complete bibliographic information, a link to the original publication on http://mededu.jmir.org/, as well as this copyright and license information must be included.
Professional isolation is an important factor in low rural health workforce retention.
The aim of this study was to gain insights to inform the development of an implementation plan for a virtual community of practice (VCoP) for general practice (GP) training in regional Australia. The study also aimed to assess the applicability of the findings of an existing framework in developing this plan. This included ascertaining the main drivers of usage, or usefulness, of the VCoP for users and establishing the different priorities between user groups.
A survey study, based on the seven-step health VCoP framework, was conducted with general practice supervisors and registrars—133 usable responses; 40% estimated response rate. Data was analyzed using the t test and the chi-square test for comparisons between groups. Factor analysis and generalized linear regression modeling were used to ascertain factors which may independently predict intention to use the VCoP.
In establishing a VCoP, facilitation was seen as important. Regarding stakeholders, the GP training provider was an important sponsor. Factor analysis showed a single goal of usefulness. Registrars had a higher intention to use the VCoP (
Important factors for a GP training VCoP include the following: facilitation covering administration and expertise, the perceived usefulness of the community, focusing usefulness around knowledge sharing, and overcoming professional isolation with high-quality content. Knowledge needs of different users should be acknowledged and help can be provided online, but trust is better built face-to-face. In conclusion, the findings of the health framework for VCoPs are relevant when developing an implementation plan for a VCoP for GP training. The main driver of success for a GP training VCoP is the perception of its usefulness by participants. Overcoming professional isolation for GP registrars using a VCoP has implications for training and retention of health workers in rural areas.
Professional isolation is an important factor in low rural health workforce retention [
Communities of practice (CoPs) are “groups of people who share a concern or a passion for something they do and learn how to do it better as they interact regularly” [
More recently, online technology has been enabling medical information sharing on an unprecedented scale [
In this context, two studies have shown that there is the interest, ability, and Internet access among general practice (GP) registrars and supervisors to establish a VCoP for GP training in a regional area of New South Wales, Australia [
The aim of this study was to gain insights to inform the development of an implementation plan for the Virtual Community of Practice (VCoP) for General Practice Training in regional Australia. The VCoP platform was based on the NING social networking software [
Ethics approval was obtained from the University of Wollongong’s Human Research Ethics Committee.
The sampling frame comprised all general practice registrars, supervisors, and educators in Coast City Country General Practice Training (CCCGPT). CCCGPT provides general practice training in a 160,000 square kilometer region of Australia, covering urban, regional, and small rural centers in the Australian Capital Territory and New South Wales. After 2 hospital years, GP registrars progress through a minimum of three general practice terms of 6 months.
In October 2011, an email was sent to the GP training provider database by the training provider administration, inviting recipients to fill in an online survey. The GP training provider database keeps an accurate record of registrars and their email addresses, listed by date. The registrar sampling frame was 143. The supervisor database is less accurate as supervisors’ details are not updated each term, while registrars’ details are. Supervisor emails are not always updated when they change and there is no date range to retrospectively check when they were active as supervisors or having a break from training. Given these limitations, a manual review by the training provider administration of the list of supervisors within the training program, cross-checked against the training program database, gave a supervisor sampling frame of 175, giving a total registrar and supervisor sampling frame of 318. In the invitation email, there was a link to SurveyMonkey, a Web-based survey program (SurveyMonkey, LLC, Palo Alto, CA, USA), with a survey and participant information sheet. A total of 183 out of 318 people responded, yielding a 57.5% response rate; 50 cases were removed for not providing consent or demographics (n=12) or for not completing the majority of the survey (n=38). Some of these noncompletions were due to emails going to practice management staff rather than doctors. The total usable response rate was 41.8% (133/318): registrar 46.9% (67/143) and supervisor 37.7% (66/175).
The questionnaire was based on previous studies demonstrating GP registrar and supervisor interest in a VCoP, and a framework that guides the implementation of health VCoPs [
The seven steps of the health VCoP framework are as follows: (1) organizing facilitation; (2) engaging stakeholders; (3) establishing clear goals; (4) involving a broad church of participants; (5) creating a supportive environment; (6) including measurement, benchmarking, and feedback in the design; and (7) technology and community factors, such as users self-selecting and having a mixture of face-to-face and online engagements. There were 28 questions in the final survey. Questions included categorical and 5-point Likert scale response items. The questions collected information on each of the seven steps, to investigate whether the steps were applicable to a VCoP for GP training. This included questions in which respondents rated the importance of a step, along with questions seeking further detail on that step to help guide the VCoP implementation. In addition, questions were asked to assess the knowledge needs of registrars when implementing guidelines, so that information on the appropriate content for the site could be obtained. Items about the features of the site were included to determine which tools would be most useful. The survey is included as
The instrument was piloted with 2 GP registrars, 2 supervisors, and 4 researchers. Discussion among this group led to some minor alterations to clarify wording. Results are presented under the seven headings of the health VCoP framework.
Data were analyzed using SPSS version 19 (IBM Corp, Armonk, NY, USA). For comparison between groups, respondents were categorized as either registrar or supervisor;
Principal axis factor analysis using varimax rotation was used to determine which Likert scale items grouped naturally in questions with multiple Likert scale items; for example, the question on the practical outcomes, or usefulness, that an online network would have for that user. If eigenvalues were >1.0, factors were included. To test for the agreement between the Likert scale items, such as the five factors perceived as
General linear regression modeling was used to test the multivariate associations of independent variables such as age, training stage, and usefulness, and the dependent variable of intention to actively use an online network for GP training.
There were 133 medical practitioners in the final sample. Of these, 51.9% (69/133) were male, 57.1% (76/133) were from a rural setting, and 50.4% (67/133) were registrars. Registrars were younger (mean 36.70 years, SD 6.85) than supervisors (mean 52.62 years, SD 7.90;
To determine which questions in the survey naturally clustered together, principal axis factor analysis using varimax rotation factor analysis was performed on two groups of questions. Participants were asked these questions to verify applicability of
The first question contained five items. Participants were asked what practical outcome, or
Secondly, participants were asked about their intention to use an online network for training by ranking their likelihood of participating through reading, sharing knowledge by answering questions, and uploading new topics. The rating scale ranged from 1 (not likely) through to 5 (highly likely).
Facilitators promote engagement and maintain community standards. [
Respondents (N=132) rated the need for formal facilitation between
Preferred leader/facilitator for the online training network (multiple responses allowed).
Facilitator | Registrar (n=67), n (%) | Supervisor (n=66), n (%) | Significance (chi-square test) |
Topic expert | 33 (49) | 20 (30) | Yes: |
GParegistrar liaison | 14 (21) | 8 (12) | No |
GP supervisor | 14 (21) | 12 (18) | No |
GP training provider administrator | 19 (28) | 24 (36) | No |
Network developer (ITb) | 5 (7) | 5 (8) | No |
Network designer (doctor) | 18 (27) | 17 (26) | No |
aGP: general practice.
bIT: information technology.
There were nine comments in the
The network needs to have an initial stakeholder champion, with stakeholder support. [
Respondents (N=130) rated the need for formal support from the main stakeholder, the GP training provider, between
The importance of the GP training provider’s support was also reflected in the previous step (see
Clear objectives provide members with responsibilities and motivate them to contribute more actively. [
Participants were asked about a range of goals for the network, and the key goal, as identified by factor analysis discussed above, was
From the health VCoP framework, clear goals are supposed to encourage active participation. Registrars (mean 3.00, SD 1.14) were more likely to state that they would participate actively than did supervisors (mean 2.52, SD 0.87;
A multivariate generalized linear regression model was developed using
Consider involving different, overlapping but not competing, professional groups, different organisations and external experts. However make sure the church is not too broad... [
Respondents were supportive of a broad church of participants. The inclusion of all medical clinicians within the training provider, including GP registrars, supervisors, and medical educators, was highly supported (see
There was also much less support for participation from groups outside the training provider, including specialists, students, academics, allied health professionals, and external registrars (see
Percentage of respondents supporting participant involvement within training provider. GP: general practice.
Percentage of respondents supporting participants outside training provider.
Health VCoPs should promote a supportive and positive culture that is both safe for members, and encouraging of participation. [
Respondents were asked about the aspects that would keep them participating in an online network, including content quality, strength of relationships, financial rewards, continuing education points, and an online points system. Respondents rated the quality of online content as their first preference (mean 4.20, SD 0.63), and their second preference was the strength of the online interaction (mean 3.98, SD 0.73). The preferences in both the registrar and supervisor groups were the same.
Health VCoPs should consider measurement as a factor in their design, including benchmarking and feedback.[
Receiving emails from the community, such as comments, updates, and responses to posts, is termed
When asked how often respondents would like to be notified that another member had added information, registrars wanted notifications more frequently than supervisors. As shown in
Registrars also wanted more frequent notification than supervisors for comments being made on a topic that they had posted. Around half (34/67, 51%) of the registrars wanted to be notified every time a comment was made, compared with only 40% (26/65) of supervisors.
Online CoPs should ensure ease of use and access, along with asynchronous communication. Other options including chat and meetings can also be considered, along with the need for training. Communities are more likely to share knowledge when there is a mixture of online and face-to-face meetings, members self-select, and both passive and active users are encouraged. [
Communities of practice rely on experts and novices sharing knowledge. Respondents were asked how comfortable they were sharing their knowledge. Registrars and supervisors were both comfortable sharing knowledge with colleagues in the training program, although
VCoP research states that knowledge sharing is best achieved by a mixture of face-to-face and online interaction [
Comparisons between registrars and supervisors on notifications from the site.
Question | Registrar, n (%) | Supervisor, n (%) | |
Every day | 3 (5) | 10 (15) | |
3-4 times/week | 12 (18) | 8 (12) | |
1-2 times/week | 29 (43) | 20 (30) | |
Fortnightly | 17 (25) | 16 (24) | |
Monthly | 5 (8) | 21 (32) | |
Yes | 34 (51) | 26 (40) | |
No | 12 (18) | 24 (37) | |
Not sure | 21 (31) | 15 (23) |
Preference for site-related material, support, and knowledge (N=133).
Question | n (%) | |
Purely face-to-face | 20 (15.0) | |
Purely online | 9 (6.8) | |
A mixture of online and face-to-face | 104 (78.3) | |
Purely face-to-face | 17 (12.8) | |
Purely online | 6 (4.5) | |
A mixture of online and face-to-face | 110 (82.7) |
Building trust is also important for sharing knowledge. Respondents indicated they were significantly more likely to build trust with other members of their knowledge-sharing community through face-to-face interaction (see
In terms of the technology used, the most popular feature was shared documents and guidelines, followed by general discussion forums, private subdiscussion groups, email mailing list, videoconferencing, and lastly live chat (see
Preferred methods of building trust and receiving training.
Preferred methods | Mean (SD)a | |
Face-to-face (N=132) | 4.19 (0.67) | |
Online (N=133) | 3.69 (0.80) | |
Formal face-to-face training | 2.92 (1.33) | |
Online help (text and images) | 3.59 (1.09) |
aLikert scale ranges from 1 (not important) to 5 (very important).
Most desirable features for an online training network (N=133).
Preferred features for the technology | Responses, n | Mean (SD) |
Shared documents | 131 | 4.01 (0.82) |
Discussion forum (all) | 132 | 3.52 (0.97) |
Discussions (private) | 130 | 3.22 (1.01) |
Email listservs | 130 | 3.10 (1.13) |
Videoconferencing | 131 | 2.90 (1.17) |
Live chat | 131 | 2.47 (1.19) |
When asked for preferences on site usernames, the most popular choice was to use their own name followed by a choice of pseudonym or real name, then using a pseudonym only (see
Preference for usernames and passwords for an online training network (N=133).
Preferences for usernames and passwords | n (%) | |
Own name | 65 (48.9) | |
Pseudonym | 5 (3.8) | |
A choice | 63 (47.7) | |
Yes | 120 (90.2) |
Finally, to further examine the knowledge-sharing needs of registrars and supervisors, participants were asked about the perceived knowledge needs of registrars. The topics covered 14 broad areas of the curriculum for the first 6 months of GP training. Respondents were asked to rate each topic according to how much help GP registrars needed, firstly, in knowing guidelines and, secondly, in implementing guidelines.
Both groups agreed that registrars needed help with their knowledge of topics, but on a combined measure, supervisors felt more strongly that registrars needed help than did the registrars (see
Knowledge of topic areas covered in the first 6 months of general practice training in Australia.
Support needed for registrar learning | Registrars, |
Supervisors, |
Mean difference | |
Need help with knowledge | 3.54 (0.80) | 4.37 (0.83) | 0.83 | .01 |
Need help with implementation | 3.59 (0.81) | 4.29 (0.50) | 0.70 | <.001 |
Difference between perceived help needed by registrars for knowledge support versus support for implementation of knowledge.
Support needed for registrar learning | Mean (SD) | Mean difference (SD) | |
Need help with knowledge | 4.02 (0.76) | 0.10 (0.44) | <.001 |
Need help with implementation | 3.92 (0.80) |
When looking at the scores of the 14 individual topics, supervisors and registrars rated the importance of topics differently. For example, supervisors gave
From these results, it is evident that the findings of the health VCoP framework [
The survey results were supportive of a facilitator for the network, in particular a topic expert. The importance of a facilitator is in keeping with previous literature reviews [
However, topic expertise is not the only desirable attribute in facilitators. In
The establishment of clear goals for a VCoP is seen as an important motivator for uptake [
The importance of perceptions of usefulness as drivers of intention to use is consistent with the technology acceptance model, in which uptake of a technology is driven by its perceived usefulness, and usefulness as a driver is even stronger than ease of use [
Therefore, in the establishment of a VCoP for GP training it will be important to focus on the usefulness for supervisors and registrars. Supervisors may need more convincing about the usefulness of the VCoP than registrars and in fact the VCoP may ultimately be more useful for registrars than supervisors. However, promoting the perception of usefulness to the potential participants may encourage uptake. The perceived usefulness will rest on clear goals of improved support for knowledge sharing and overcoming professional isolation. It may even be that supervisor perception of usefulness could increase if registrars use such a VCoP and find that it achieves these goals.
A broad church of users is acknowledged as an important factor for success from the literature [
According to the health VCoP framework, another important aspect of a VCoP for health is
The findings of the health framework for VCoPs are relevant when developing an implementation plan for a VCoP for GP training. The implementation plan should involve following the seven steps of facilitation: stakeholder engagement, developing clear goals, engaging a
The sharing of knowledge to overcome professional isolation and improve connectedness is a useful goal for a VCoP. GP training can be isolating, leading to issues of workforce retention in rural areas. If professional isolation can be overcome, this may assist with the training and ultimately the retention of rural and regional general practitioners. This has broader implications beyond the training of rural general practitioners in Australia; this may inform training of medical specialists and allied health professionals as they rotate through regional placements, both in Australia and in other countries attempting to train and retain health professionals across a wide geography.
There are a number of limitations to this study. Firstly, the study was conducted in a single regional training provider. This may introduce bias around demographics and geography which could limit the generalizability of the findings. However, in terms of rural and urban comparisons, the study participants were evenly distributed across rural and urban areas, with no significant differences found based on rurality, so this may improve the confidence in the external validity of the studies.
Secondly, the response rate for the surveys was 40%, but the overall numbers were modest. Response rates to physician surveys are often lower than those for nonphysicians, but the response rate here is still a little lower than the 40-50% quoted in a review of physician response rates [
There may be a self-selection bias in this study, as it was a study regarding online attitudes and the survey was distributed via email with a survey link. This online distribution method may have encouraged responses from users with higher baseline levels of confidence with online communication. In spite of this potential bias, the overall levels of confidence and usage were at least in keeping with, if not below, the levels found in some comparative studies, such as a recent study on social media usage among physicians in the United States [
Finally, it is important to acknowledge the dynamic nature of technology. Since this survey was conducted in October 2011, new versions of technology tools have been developed with increasing functionality. The technology knowledge and skills of medical practitioners has also evolved during this period. The finding should, therefore, be read with this in mind. Despite this, the tools discussed in this study remain the foundation of many online interactions and the conceptual model we discuss can be applied to any set of technologies.
Study survey given to participants.
Coast City Country General Practice Training
community of practice
general practice
information technology
virtual community of practice
The authors would like to thank Coast City Country General Practice Training for funding and support of this study, Professor Andrew Bonney for his insights and encouragement, and Mr Lance Barrie for project support.
SB (Stephen Barnett) designed the survey and carried out the majority of the analysis and writing. SJ assisted with survey design, analysis, reviews and amendments to the paper, and final approval. SB (Sue Bennett) assisted with survey design and drafts of the paper. DI assisted with survey design, analysis, and amendments to the paper. LR performed initial data analysis and was involved in drafting the manuscript. All authors reviewed and approved the final manuscript.
Funding was provided for this study by Coast City Country General Practice Training. The first author (SB) is the Medical Director of e-healthspace, an online community for doctors.