- Research article
- Open Access
A social network analysis on clinical education of diabetic foot
© Shokoohi et al.; licensee BioMed Central Ltd. 2013
- Received: 27 March 2013
- Accepted: 12 September 2013
- Published: 20 September 2013
Identification of Educational Influentials (EIs) in clinical settings helps considerably to knowledge transfer among health and medical practice providers. The aim of this study was identifying EIs in diabetic foot ulcers (DFU) by medical students (clerks, interns and residents) and providing their relational pattern in this subject.
Subjects were medical students at clerk, intern and resident levels in a local educational hospital. A standard questionnaire with four domains (knowledge, communication, participation and professional ethics) was used for identifying EIs. Students introduced those people with these characteristics who referred them for DFU. Respective communication networks were drawn as intra-group (such as resident-resident) and inter-group (such as intern-resident) networks and quantitative criteria of density, in-degree and out-degree centrality and reciprocity were measured.
The network density of clerks-residents (0.024) and interns-residents (0.038) were higher than clerks-attends (0.015) and interns-attends (0.05); indicating that there were more consulting interactions in former networks than the latter. Degree centrality in residents-related networks (clerks-residents = 2.3; interns-residents = 2.6) were higher than attends-related ones (clerks-attends = 1.1; interns-attends = 1.7), while they were not statistically significant. However, In-degree centralization, which indicating a degree of variance of the whole network of ingoing relationships, in attends-related networks was greater than resident-related networks.
Resident were consulted with almost as same as attends on DFU. It showed that residents were playing a remarkable role in knowledge transfer and they can be considered as EIs in this clinical setting. It seemed that the availability was the main reason for this key role.
- Knowledge Transfer
- Degree Centrality
- Social Network Analysis
- United Kingdom Prospective Diabetes Study
- Diabetic Foot Ulcer
There are more than 3 million people over 20 years old with diabetes in Iran. One of the major complications is diabetic foot ulcers (DFUs) and the sever consequence is foot amputation. Disability Adjusted Lost Years (DALY) of type 2 diabetes in Iran was estimated at 4.7 years per 1000  and burden of diabetic foot ulcers and organ amputation were assessed to 5848 and 1573 lost years, respectively . In 2002, approximately 100,000 deaths occurred in Iran due to diabetes . The United Kingdom prospective diabetes study (UKPDS) indicated that 20% of diabetic people suffer from diabetic neuropathy within 4 years after diagnosis of diabetes, which reaches to 50% after 15 years . Lavery et al. in a study on Mexican–Americans and non-Hispanic Whites found that foot amputation is 5.9 per 1,000 ones annually .
Health care providers and medical staff including physicians and nurses are the key groups, which could prevent deteriorating problems of DFUs by providing appropriate health care for the patients [6–8]. Studies in Iran have shown that the knowledge, attitude and practice of physicians in the field of diabetes care, especially in the fields of patients“ education, treatment and care of complications, was insufficient . Even studies in developed countries including USA and Netherlands have shown that 30-40% of the patients did not receive health care based on best available evidence, and about 20-25% of the provided cares were insufficient or even potentially harmful [10–12].
Educational Influential (EI) concept and its importance have been introduced to medical education few decades ago. Hiss et al. were among the first people who found the importance of this issue and developed the method for identification of EI in the medical field . Various studies have shown that these people were distinct from other colleagues due to their specific characteristics and had gained an influential and strong role in knowledge transfer and diffusion of innovations. They can lead others to the quality of care and subsequently affect the patients’ health [14–16].
Studies indicated that the most trusted and influential source of information for physicians were their colleagues who communicating with them. Relationships between individuals within a network is as an important source for clinical knowledge and decision making [17, 18]. This knowledge significantly influence decision making and management of the patients [19–21]. Pathman et al. proposed a process for implementation of guidelines including awareness, intellectual agreement, decision, adoption and adherence to the guideline or its integration in the routine practice . Traditional continuous medical education (CME) programs generally stopped in the level of awareness or agreement so that they did not reach to the adopt stage. Thus, nowadays considering the adoption stage is an important target in the CMEs. EIs are those people who influence on others’ opinions, attitudes, beliefs, motivations, judgments and behaviors. They can affect adoption of innovation by other practitioners. Knowledge transfer would be easier by identification of these people since in this way, knowledge reached more effectively to the recipients .
One way to determine influential individuals within a network, e.g. a hospital, is the social network analysis (SNA). The interpersonal relationships and knowledge transfer patterns within networks can be investigated by using this analysis. Social network consists of a set of actors and their relationships . Tindall and Wellman defined the SNA as analyzing the social structure and its impacts . This analysis can be used for studying the influence of networks’ structure on physicians’ attributes, dissemination of medical knowledge and implementation of the guidelines [26–28].
The aim of this study was identifying the health care providers’ social network in a sample hospital in Iran and knowledge transfer relationships regarding DFU. The specific objective was comparing the role of attends and residents in the aforementioned network.
This study was conducted in one of the educational hospitals with about 400 beds in one of the medical sciences universities in Iran. The hospital had five wards including general, surgery, pediatrics, obstetrics and dermatology with approximately 80 faculty members. 140 students participated in this study including 70 clerks, 45 interns and 25 residents.
Students were asked to introduce whom they were consulted with for solving their problem (questions) on DFU care. Hiss et al. developed a sociometric instrument for identifying EIs for the first time. It included three domains: knowledge ability, communication skills and humanism. However, authors of the present study developed an identical sociometric questionnaire which published elsewhere and compared characteristics of EI in developed and developing countries . In the later questionnaire, which was used in the present study, the EIs can be identified by four criteria including high level of knowledge, communication skills, taking into account of the stakeholders and following professional ethics.
In the present study, the network was assumed as egocentric. It means that when density is to be calculated, ego nodes and their direct relations are ignored and only relationships between other nodes are considered. Four groups of indicators were estimated in the present study including density, centralization (including in-degree and out-degree), degree centrality and Reciprocity (including dyad and arc).
It is a description of general situation of links or interconnection of points in the network . This indicator shows the ratio of the existing relations between individuals to a maximum number of possible relationships between them which is measured as a value between zero and one. In a network, the more actors exist in the relationship; the network density will be higher [31–33]. Network density may show the information flow among the nodes.
Centralization implies influential people within a network. Network centralization score ranged between zero to one, that is, zero means every network individual have relationship with every other member, and one means that all members are connected to only one individual. In the other words, the centralization implies the degree of asymmetry in the network. When some individuals have more connections than others, it is expected to have a high centralization score. Centralization has two sub measurements in a directed network: 1) In-degree centralization: degree of variance of the total network of ingoing relationships in comparison with an ideal star network (which, theoretically, implies the most centralization). Centralization provides an analysis of centrality of consulted individuals. If it is close to one, it means that few individuals of the network are consulted by the rest of the individuals. 2) Out-degree centralization: degree of variance of the total network of outgoing relationships in comparison with an ideal star network. It shows consulted individuals’ situation, and its high level indicates that small individuals of the network do most of consulting of others [31, 33].
The mean number of (in/out) going relationships of a person. People who receive many links may be prominent, while subjects who have many connections to others may be influential. The measure refers to direct connections to an individual only .
Which indicating relationship or interaction between the two individuals symmetrically, that is, if actor A asks consultation from actor B, if actor B asks consultation from actor A too . This indicator was calculated in two ways: 1) dyad (or hybrid); which means there are several pairs or reciprocated relationship between the actors. This method is obtained from the number of pairs of communication between two individuals within the network divided by existing communications in the network. 2) The arc is the other one; i.e. instead of focusing on the number of existing communications, it is focused on the maximum possible number of communications within the network, by focusing on all possible communications within the network.
All of these indicators were calculated for three kinds of communication networks in the following subgroups: A) Intergroup Communication Networks: including six subgroups on 1. Clerks with interns, 2. Clerks with residents, 3. Clerks with attends, 4. Interns with residents, 5. Night interns with attends, and 6. Residents with attends. B) Intra-group Communication Networks: including three subgroups on 1. Clerks with clerks, 2. Interns with interns, and 3. Residents with residents.
UCInet 6 software for windows was used for determining the centrality criteria, providing maps for indicating interpersonal relationships and visual identification of EIs as well as data analysis. NetDraw 2.41 (Network Visualization Software) (a UCInet subprogram) was used for drawing network diagrams [35, 36].
The present study was approved by the ethics committee of Tehran University of Medical Sciences (TUMS). An informed consent was obtained from all students who participated to the study.
Description of inter-subgroup networks
Inter subgroups networks
Clerks with interns
Clerks with residents
Clerks with attends
Interns with residents
Interns with attends
Residents with attends
In-degree centralization of network*
In-degree centrality, mean ± SD (min, max)
0.58 ± 1.74 (0, 9) *
2.3 ± 5.9 (0, 32)
1.1 ± 4.9 (0, 32)
2.6 ± 5.9 (0, 29)
1.7 ± 4.5 (0, 27)
3.3 ± 4.1 (0, 16)
Out-degree centralization of network
Out-degree centrality, mean ± SD (min, max)
0.58 ± 0.65 (0, 2)
2.3 ± 0.73 (0, 4)
1.1 ± 0.53 (0, 2)
2.6 ± 0.89 (0, 4)
1.7 ± 0.93 (0, 3)
3.3 ± 1.5 (0, 5)
0.005 ± 0.07
0.024 ± 0.15
0.015 ± 0.12
0.038 ± 0.19
0.05 ± 0.18
0.11 ± 0.31
The mean degree centrality in resident-related networks was almost higher than attends-related ones. The mean degree centrality for clerks-residents and interns-residents was 2.3 and 2.6, while for clerks-attends and interns-attends was 1.1 and 1.7, respectively. The maximum degree centrality belonged to resident-attends network, while the minimum was to clerks-interns network.
The mean network density of clerks-residents (0.024) and interns-residents (0.038) were higher than that in clerks-attends (0.015) and interns-attends (0.05) networks, indicating that there were more consulting interactions in former networks than the latter. The maximum mean density was related to residents-attends network (0.11 ± 0.31). The reciprocity indicator was zero for all of the subgroups based on both of dyad and arc methods (not showed in the table).
Description of intra-subgroup networks
Intra subgroups networks
Residents with residents
Interns with interns
Clerks with clerks
In-degree centralization of network (%)
In-degree centrality (Mean ± SD) (min, max)
2.2 ± 3.3 (0, 13)
0.57 ± 1.2 (0, 5)
0.0 ± 0.0 (0, 0)
Out-degree centralization of network (%)
Out-degree centrality (Mean ± SD) (min, max)
2.2 ± 0.98 (0, 4)
0.57 ± 0.69 (0, 2)
0.0 ± 0.0 (0, 0)
0.091 ± 0.28
0.013 ± 0.11
0.0 ± 0.0
Dyad or hybrid
This study indicated that the communication degree and the distribution of connections in residents-related networks and attends- related ones were nearly similar to each other. In the other words, not only the residents were in relationship with each other, but also the other subgroups including interns and clerks communicated with the residents. This relationship was nearly similar to the communications that the clerks and interns had with their attending.
There are four approaches for identifying EIs: sociometric methods, key informant methods, self-designing methods and observation methods. Most of the studies for identification of EIs used the sociometric instrument, which we used in the present study. A recent systematic review showed that nine trails out of twelve clinical trials used this method .
Relatively low density of these networks may be attributed to the type of network, because when density was measured in egocentric method (unlike density measurement based on socio-centric method), ego nodes (those who were consulted with and communicated with other nodes) were not considered. This characteristic of ego-centric network was regarded as a privilege in SNA methods since it sought for finding relationship of other network individuals with each other. In a qualitative study aiming at identification of obstacles in using evidence based medicine and fail to answer their clinical questions, it was shown that several factors including shortage of enough time, lack of access to electronic resources, not having skills required to search electronic resources, lack of tracking unanswered questions, lack of priority for some questions and lack of personal initiative in tracking questions were the barriers. More importantly, team dynamics and institutional culture were considered as the main obstacles. It was indicated that dynamism or team work among medical students especially residents and hospital dominant culture could considerably help transfer of knowledge . In the present study, one reason for justifying low density in communication subgroups, especially in resident - attend and resident - resident subgroups, might be due to the fact that this group of students searched sources other than their colleagues (books, papers, …) for finding the answers for their questions related to the complicated cases such as DFU.
Rappolt performed a qualitative study on 45 family physicians in order to identify sources of clinical information gathering and found that the first source was informal consultants with peers (44%) due to easy accessibility and approachability of them. The second and the third sources for gathering clinical information were asking peer experts (24%) and searching related literature (22%), respectively . In addition, it could be mentioned that the responsibility burden on residents was often higher than the other groups which could justify high communication of this group with others in comparison to other networks.
Reciprocity could show communications among the residents. According to arc indicator, it was 7.2%, implying interaction between the individuals within resident - resident network. However, this indicator was zero for intern - intern and clerk - clerk subgroups, showing lack of interaction among these subgroups. It was clear that the later groups were at the initial states of service provision and possible clinical stresses did not allow them to take risk and rely on each other. However, the residents, who were at higher levels of clinical decision making, consulted with each other. Thus, experience can be considered as an important factor in seeking for consultation from colleagues.
In a survey study conducted at a teaching hospital in Boston, Keating et al. showed that the clinical information had a directed from highly experienced colleagues and accessible colleagues based on the location and schedule. Other predictors to take the clinical consultations among those physicians was having the same gender .
Considering in-degree centrality in clerk - attend, clerk - resident, intern - resident, and intern - attend, resident, it is clear that not only the attends could be the consultation source, but also the residents played the role of EI in this setting. High values for these indicators, whether in average or in the range level, suggested the existence of influential individuals within the related networks.
Flodgren et al. in a systematic review found that EIs independently or in conjunction with other interventions could improve evidence-based practice successfully. Effectiveness varied within and between the studies due to high heterogeneities of the collected studies. Therefore, there was no clear way to show and optimize the effectiveness of EIs .
Network approaches and science can provide some concepts and techniques to scrutinize the connections between people in any system. Although, network approaches’ application in medical care research is fairly new, in many scientific disciplines such as neurosciences, molecular life sciences, and public health, it has been used. Recent research including studies of patient care teams in treatment of diabetes and chronic heart failure in primary care , opinion networks of long-term care specialists , and connectedness of health care professionals in the treatment of Parkinson . A social network approach may be particularly relevant if actors (people) have flawed knowledge on their performance options and the disease-related outcomes.
The authors would like to thank medical students (clerks, interns and residents) to participate in the present study. A part of the present study was the Master’s thesis of the first author (MS) and was financially supported by the Tehran University of Medical Sciences.
- Abol Hassani F, Mohajeri Tehrani MR, Tabatabaei O, Larijani B: Burden of diabetes and its complications in Iran in year 2000. Iran J Diabetes and Lipid Disord 2005, 5: 35–48.Google Scholar
- Naghavi M: A Study on Burden of Disease, Accidents, and Risk Factors in the Islamic Republic of Iran. Tehran: Ministry of Health and Medical Education; 2008.Google Scholar
- Delavari AR: National Program for Control and Prevention of Diabetes. Tehran: Center for Disease Control; 2004.Google Scholar
- Tight blood pressure control and risk of macrovascular and microvascular complications in type 2 diabetes: UKPDS 38: UK Prospective Diabetes Study Group. Bmj 1998, 317: 703–713.View ArticleGoogle Scholar
- Lavery LA, Armstrong DG, Wunderlich RP, Tredwell J, Boulton AJM: Diabetic foot syndrome: Evaluating the prevalence and incidence of foot pathology in Mexican Americans and Non-Hispanic Whites from a diabetes disease management cohort. Diabetes Care 2003, 26: 1435–1438.View ArticlePubMedGoogle Scholar
- Aalaa M, Tabatabaei Malazy O, Sanjari M, Peimani M, Mohajeri-Tehrani MR: Nurses’ role in diabetic foot prevention and care; a review. J Diabetes Metab Disord 2012, 11: 2–6.View ArticleGoogle Scholar
- Aliasgharpour M, Dehghan-Nayeri N: The care process of diabetic foot ulcer patients: a qualitative study in Iran. J Diabetes Metab Disord 2012, 11: 2–9.View ArticleGoogle Scholar
- Madanchi NT-MO, Pajouhi M, Heshmat R, Larijani B, Mohajeri-Tehrani MR: Who are diabetic foot patients? A descriptive study on 873 patients. J Diabetes Metab Disord 2013, 12: 2–6.View ArticleGoogle Scholar
- Peimani M, Tabatabaei-Malazy O, Heshmat R, Amiri Moghaddam S, Sanjari M, Pajouhi M: Knowledge, Attitude and Practice of physicians in the field of diabetes and its complications; A pilot study. Iranian Journal of Diabetes and Lipid Disorders 2010, 9: 1–7.Google Scholar
- Berardis G, Pellegrini F, Franciosi M, Belfiglio M, Di Nardo B, Greenfield S, et al.: Are Type 2 diabetic patients offered adequate foot care? The role of physician and patient characteristics. J Diabetes Complications 2005, 19: 319–327.View ArticlePubMedGoogle Scholar
- Litaker D, Flocke SA, Frolkis JP, Stange KC: Physicians’ attitudes and preventive care delivery: insights from the DOPC study. Prev Med 2005, 40: 556–563.View ArticlePubMedGoogle Scholar
- Mason J, O’Keeffet C, Mclntosht A, Hutchinson A, Booth A, Young RJ: A systematic review of foot ulcer in patients with type 2 diabetes mellitus. I: Prevention. Diabetic Medicine 1999, 16: 801–812.PubMedGoogle Scholar
- Hiss RG, Macdonald R, Davis WK: Identification of physician educational influentials (EI's) in small community hospitals. In Proceedings of the Seventeenth Annual Conference on Research in Medical Eduaction. Washington, DC; 1978:283–288.Google Scholar
- Lippet R, Watson J, Westley B: The Dynamic of Planned Change. New York: Harcourt, Brace and World, Inc; 1958.Google Scholar
- Wenrick JW, Mann FC, Morris W, Reilly A: Informal Education for Practicing Physicians. J Med Educ 1971, 46: 299–305.Google Scholar
- Wright FC, Ryan DP, Dodge JE, Last LD, Law CH: Identifying educationally influential specialists: issues arising from the use of “classic” criteria. J Contin Educ Health Prof 2004, 24: 213–226.View ArticlePubMedGoogle Scholar
- Rappolt S: Family physicians’ selection of informal peer consultants: implications for continuing education. J Contin Educ Health Prof 2002, 22: 113–120.View ArticlePubMedGoogle Scholar
- Rogers EM: Diffusion of innovations. New York: Free Press; 1995.Google Scholar
- Covell DG, Uman GC, Manning PR: Information needs in office practice: are they being met? Ann Intern Med 1985, 103: 596–599.View ArticlePubMedGoogle Scholar
- Cullen R: The medical specialist: information gateway or gatekeeper for the family practitioner. Bull Med Libr Assoc 1997, 85: 348–355.PubMed CentralPubMedGoogle Scholar
- Williamson JW, German PS, Weiss R, Skinner EA, Bowes F 3rd: Health science information management and continuing education of physicians. A survey of U.S. primary care practitioners and their opinion leaders. Ann Intern Med 1989, 110: 151–160.View ArticlePubMedGoogle Scholar
- Pathman DE, Konrad TR, Freed GL, Freeman VA, Kock GG: The awareness – to- adherence model of the steps to clinical guideline compliance: the case of pediatric Vacine Recommendations. Med Care 1996, 34: 873–889.View ArticlePubMedGoogle Scholar
- Nissen ME: An extended model of knowledge-flow dynamics. Commun Assoc Inf Syst 2002, 8: 251–266.Google Scholar
- Hanneman R, Riddle M: Introduction to social network methods. Riverside, CA: University of California, Riverside; 2005. (published in digital form at ) http://faculty.ucr.edu/~hanneman/nettext/Introduction_to_Social_Network_Methods.pdfGoogle Scholar
- Tindall D, Wellman B: Canada as social structure: Social network analysis and Canadian sociology. Can J Sociol 2001, 26: 265–308.View ArticleGoogle Scholar
- Fattore G, Frosini F, Salvatore D, Tozzi V: Social network analysis in primary care: The impact of interactions on prescribing behaviour. Health Policy 2009, 92: 141–148.View ArticlePubMedGoogle Scholar
- Gifford D, Holloway R, Frankel M, Albright CL, Meyerson R, Griggs RC, Vickrey BG: Improving adherence to dementia guidelines through education and opinion leaders. A randomized, controlled trial. Ann Intern Med 1999, 131: 237–246.View ArticlePubMedGoogle Scholar
- Soumerai SB, McLaughlin TJ, Gurwitz JH, Guadagnoli E, Hauptman PJ, Borbas C, et al.: Effect of local medical opinion leaders on quality of care for acute myocardial infarction: a randomized controlled trial. JAMA 1998, 279: 1358–1363.View ArticlePubMedGoogle Scholar
- Shokoohi M, Nedjat S, Golestan B, Soltani A, Majdzadeh R: Can Criteria for Identifying Educational Influentials in Developed Countries be Applied to Other Countries? A Study in Iran. J Contin Educ Health Prof 2011, 31: 95–102.View ArticlePubMedGoogle Scholar
- Freeman LC: Centrality in social networks I: conceptual clarifi cation. Soc Networks 1979, 1: 215–239.View ArticleGoogle Scholar
- Scott JTA, Crosson JC, Orzano AJ, Stroebel C, DiCicco-Bloom B: Social Network Analysis as an Analytic Tool for Interaction Patterns in Primary Care Practices. Ann Fam Med 2005, 3: 443–448.PubMed CentralView ArticlePubMedGoogle Scholar
- Weenink JW, van Lieshout J, Jung HP, Wensing M: Patient Care Teams in treatment of diabetes and chronic heart failure in primary care: an observational networks study. Implement Sci 2011, 6: 2–9.View ArticleGoogle Scholar
- Wensing M, van der Eijk M, Koetsenruijter J, Bloem BR, Munneke M, Faber M: Connectedness of healthcare professionals involved in the treatment of patients with Parkinson’s disease: a social networks study. Implement Sci 2011, 6: 67.PubMed CentralView ArticlePubMedGoogle Scholar
- Barbara Z, Hlebec V: Reliability of Measures of Centrality and Prominence. Soc Networks 2005, 27: 73–88.View ArticleGoogle Scholar
- Borgatti SP, Everett MG, Freeman LC: Ucinet for Windows: Software for Social Network Analysis. Harvard, MA: Analytic Technologies; 2002.Google Scholar
- Borgatti SP, Foster PC: The network paradigm in organizational research: a review and typology. J anagement 2003, 29: 991–1013.Google Scholar
- Doumit G, Gattellari M, Grimshaw J, O’Brien MA: Local opinion leaders: effects on professional practice and health care outcomes. Cochrane Database Syst Rev 2007,24(1):CD000125. 39Google Scholar
- Green ML, Ruff TR: Why Do Residents Fail to Answer Their Clinical Questions? A Qualitative Study of Barriers to Practicing Evidence-Based Medicine. Acad Med 2005, 80: 176–182.View ArticlePubMedGoogle Scholar
- Keating NL, Ayanian JZ, Cleary PD, Marsden PV: Factors Affecting Influential Discussions among Physicians: A Social Network Analysis of a Primary Care Practice. J Gen Intern Med 2007, 22: 794–798.PubMed CentralView ArticlePubMedGoogle Scholar
- Flodgren G, Parmelli E, Doumit G, Gattellari M, O’Brien MA, Grimshaw J, Eccles MP: Local opinion leaders: effects on professional practice and health care outcomes. Cochrane Database Syst Rev 2007, 1: CD000125. DOI:10.1002/14651858.CD000125.pub3Google Scholar
- Clark MA, Linkletter CD, Wen X, et al.: Opinion networks among long-term care specialists. Med Care Res Rev 2010, 67: 102S-125S.View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. 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 is properly cited.