Knowledge sharing among healthcare infection preventionists: the impact of public health professionals in a rural state
© Wiemken et al.; licensee BioMed Central Ltd. 2012
Received: 27 April 2012
Accepted: 13 July 2012
Published: 28 July 2012
Healthcare-associated infections are a major source of morbidity and mortality in the United States. Infection Preventionists (IPs) are healthcare workers tasked at overseeing the prevention and control of these infections, but they may have difficulties obtaining up-to-date information, primarily in rural states. The objective of this study was to evaluate the importance of public health involvement on the knowledge-sharing network of IPs in a rural state.
A total of 95 attendees completed our survey. The addition of public health professionals increased the density of the network, reduced the number of separate components of the network, and reduced the number of key players needed to contact nearly all of the other network members. All network metrics were higher for public health professionals than for IPs without public health involvement.
The addition of public health professionals involved in healthcare infection prevention activities augmented the knowledge sharing potential of the IPs in Iowa. Rural states without public health involvement in healthcare-associated infection (HAI) prevention efforts should consider the potential benefits of adding these personnel to the public health workforce to help facilitate communication of HAI-related information.
Healthcare-associated infections (HAIs) are a major source of morbidity and mortality, causing an estimated two million infections and 100,000 deaths each year . From a public health perspective, the number of HAIs exceeds the number of cases of any notifiable disease, and the deaths associated with HAIs are greater than the number of deaths attributable to several of the top ten leading causes of death . The importance of HAIs to public health practice was highlighted by the addition of a Healthy People 2020 goal to “prevent, reduce, and ultimately eliminate healthcare-associated infections (HAIs) ” . The steady increase in the number of states with legislative mandates for the reporting of HAIs has increased the need for communication between healthcare, public health, and government in terms of identifying HAIs, reporting HAIs, and improving patient outcomes.
Existing relationships between healthcare facility infection preventionists (IPs) and their public health colleagues varies greatly from state to state. States such as New York, California, and Pennsylvania have sophisticated approaches to collaboration that have resulted in advances in prevention efforts [3, 4]. In rural states, where a substantial number of hospitals are small critical access hospitals, public health departments and IPs are confronted with unique challenges. For example, with the exception of a few studies, [5–9] basic epidemiologic data on HAIs in rural states are missing from the literature. These lack of data prevent many targeted interventions from ever being realized. Furthermore, rural states may have limited resources compared to more metropolitan states, and therefore may be forced to leverage existing relationships to implement interventions that rely heavily upon their current capabilities. The Commonwealth of Kentucky provides an example of such a rural state.
Kentucky has a population of approximately 4 million  and is 56% rural . By comparison, the US is considered to be 19% rural . A recent survey of Kentucky’s IPs revealed that 25% have less than 5 years of experience in that role,  they lack formal mentoring programs, and their job responsibilities often exceed their education and experiential preparation . In order to address this training and knowledge gap, it is critical to understand how knowledge is shared among these professionals. An analysis of the existing knowledge-sharing network in Kentucky was conducted in 2010, and found that there were distinct gaps in the ability of the IPs to gain and share knowledge with respect to infection prevention and control . One of the recommendations from this study was to define ways to increase knowledge sharing between IPs.
The state of Iowa is similar to Kentucky in that it is predominantly rural (70%) and has a similar configuration of hospital sizes . Iowa differs from Kentucky in that they have long-standing, state wide healthcare-public health partnerships. These partnerships may provide a network structure more conducive to knowledge sharing, resulting in decreased efforts necessary in conducting and coordinating infection prevention activities.
The purpose of this study was to define the role of public health professionals on the knowledge-sharing network of IPs in a rural state.
Study design and population
A social network analysis was conducted to evaluate the impact of public health professionals on the knowledge-sharing network of IPs in Iowa. A whole network approach was used . A paper-based survey was provided to all attendees at a statewide infection control meeting in Iowa in May 2011. Instructions were given orally and were included on the survey instrument. Completed surveys were collected approximately 30 minutes after initiation. The survey has been previously utilized in another communication network protocol  and was pilot tested prior to the previous study and the current study.
A member of the network with whom another member of the network shares knowledge, an IPs communication contacts [here, used only in the definition of eigenvector centrality].
A member of the network that receives communication from many hubs. This person receives information from potentially important people in the network.
How often a member of the network falls between the shortest knowledge-sharing path of two other members of the network. For example, if two IPs have a mutual contact but cannot, themselves, communicate with one another, their mutual contact serves as a broker of communication, a network node with high betweenness, as he or she falls within the shortest communication path between the other two network members.
A node connected to a diverse set of other nodes
How much the network is centered around a few central members.
A subgroup of members of the network in which each person is connected to every other member of the subgroup.
A sub-group that was not connected to any sub-group in the network.
A measure of bridging: low constraint indicates connection to others who are not themselves connected.
The relative number of knowledge-sharing episodes of a particular member of the networks alters. For example, if an IP#1 has one alter (IP#2) who themselves has one alter (IP#3), the maximum amount of knowledge that can be gained by IP#1 is, at most, the sum of the knowledge of IP#2 and IP#3 (low eigenvector centrality for IP#1). However, if IP#2 has 80 contacts, the knowledge potential through the IP#1 to IP#2 communication path is much greater since IP#2 has a lot of different communication paths with which to obtain information (high eigenvector centrality).
Fragmentation Key Player
A member of the network capable of holding the network together in as few components as possible while taking into account the size of each component as well as the other fragmentation key players. This is an IP that, if removed can break up the network into many pieces, cutting off certain members from obtaining information from the rest of the network.
A member of the network that provide information many authorities. These members provide information to potentially important members of the network.
The number of times a member of the network was asked for knowledge.
A member of the network with no connections to any other member of the network.
Having exchanged infection prevention-related information formally or informally via any method with another member of the network in Iowa over the past 6 months.
A survey respondent.
The number of times a member of the network asked another member of the network for knowledge.
Reach Key Player
A member of the network that had the capability of sharing knowledge with the largest proportion of other nodes in the network while taking into account the other reach key players and the potential for redundant connections. This is an IP that has a lot of contacts that other members of the network do not share, making them a unique individual with respect to the number of other members of the network with which they can communicate.
Network cohesion was measured by density clique, and component analyses. Node centrality was measured using in-degree out-degree betweenness, and eigenvector centrality measures. Centralization indices were used to demonstrate the level to which the network focused on one particular central node. Bridges were identified using the constraint statistic . Hub and authority scores were calculated using generalizations of eigenvector centrality . Key players in the network were identified using the key player algorithm developed by Borgotti and colleagues . Both reach and fragmentation key players were identified using sequential addition of key players. The final number of key players in each analysis was identified when the reach index or fragmentation index did not increase more than 10% with the addition of the extra key player . The multiple group sizes algorithm was used to compute the total number of key players necessary to reach 100% of the other members of the main component of the network. UCINET and NETDRAW were used for all social network analyses (Analytic Technologies, Lexington, KY). KeyPlayer v1 was used for identification of key players (Analytic Technologies, Lexington, KY). SAS v9.2 (SAS Inc. Cary, NC) was used for other descriptive analyses.
To evaluate the impact of the public health department coordination in Iowa, all analyses were conducted twice: once with only the IPs and once with the IPs and the health department personnel.
Human subjects protection
Institutional review board approval was granted by the University of Louisville Human Subjects Protection Office (Protocol #08.0399) prior to any data collection. Written informed consent was waived for all data collection activities, and no data were collected from children.
Demographic characteristics of 91 responding infection preventionists in Iowa
2-3 Year College (Including RN)
4 Year College
Professional Degree (ARNP, PA, etc.)
Regular APIC Meeting Attendance – Local Chapter
Regular APIC Meeting Attendance – Statewide Meeting
Time in Current Position
Less than 1 Year
More than 10 Years
Daily Efforts in Infection Prevention
More than 50% but less than 100%
Length of Career in Infection Prevention
Less than 1 Year
More than 10 Years
Affiliation with an Academic Center
Use of the National Healthcare Safety Network
Centrality and centralization statistics of the knowledge-sharing social network of hospital-based infection preventionists and public health pofessionals in Iowa
4.5 (3 – 19)
1.3 (0 – 950.5)
0.02 (0.13 – 0.32)
0.04 (0 – 0.28)
0.01 (0 – 0.47)
0.13 (0.08 – 0.57)
Cohesion and key player statistics for the knowledge-sharing social network of hospital-based infection preventionists and public health professionals in Iowa
1.7% (274 ties)
1.9% (336 ties)
n = 11
n = 9
n = 3
n = 2
n = 4
n = 4
Reach Key Players
n = 4 (72% Reachable)
n = 3 (74% Reachable)
For 100% Reach of Main Component
n = 23
n = 21
Fragmentation Key Players
n = 4 (54% Fragmentation)
n = 3 (34% Fragmentation)
Our results demonstrate that complementing the IP knowledge network with public health professionals may increase the ability of IPs to share knowledge with each other. Since many IPs in Iowa communicate with the public health department regarding HAIs those public health professions play a unique role within the structure of the Iowa network. Our findings that public health professionals in Iowa have higher network statistics suggest that these members have important structural roles within the infection prevention knowledge network. The higher in-degree statistics for public health professionals suggest that they are contacted by many IPs, while higher out-degree statistics suggest they provide information to many members of the network. The larger betweenness statistics suggest public health professionals can connect otherwise disconnected members of the network, and higher eigenvector statistics suggest that these members contact IPs who themselves contact a large number of other IPs. These findings provide evidence that information can quickly diffuse through this network when provided to public health professionals.
One important finding was that public health members are able to connect otherwise isolated IPs. Although the density of the network did not increase a great deal (0.2%) with the addition of public health professionals, they added 62 communication ties to the network, a very large number for only 4 additional members. These additional ties will allow network members to maintain communication if some IPs leave their positions. Key player metrics further demonstrated the importance of these members within the knowledge-sharing network. With the addition of the public health members, communication between nearly every member of the network was possible after initiating communication through a smaller set of key members. In the event of a pandemic or outbreak, it is critical to quickly convey new information to nearly every member of the network, and focusing communication efforts on these reach key players may provide an ideal mechanism for improving the flow of information. Finally, we found that the public health members reduced the network fragmentation upon removal of fragmentation key players. Because of the presence of these public health professionals and their unique placement within the structure of the network, the network will be less affected when other members leave the network (e.g. retirement or changing jobs).
Another important concept in communication networks is “bridging”. Members who represent bridges are capable of connecting members of a network that do not readily connect with each other. These members are considered to be personnel with critical influence for improving team performance in a network [19, 22, 23]. Here, we identified that the network member with the lowest constraint score (indicative of a good bridge) was a public health professional. This bridge allows for dissemination of information from diverse areas of the network. As various areas of the network may hold members with different areas of expertise, bridges allow for this expertise to flow to and from these areas. For example, if some IPs in a network are strong in prevention of ventilator-associated pneumonia, and another is proficient with use of the National Healthcare Safety Network (NHSN), the network bridge is capable of connecting these IPs. Bridges are particularly important in a rural state where most of the hospitals are critical access hospitals. In these facilities, IPs have multiple demands and limited resources, making it difficult to gain expertise in multiple areas. Because of these factors, it is especially important to identify bridging network members capable of gathering and disseminating a wide variety of HAI-related information.
Another possible method of increasing knowledge sharing is periodic statewide meetings of infection preventionists . However, despite the presence of a long-standing annual statewide IP meeting in Iowa, there were many similarities between the Iowa and the Kentucky IP knowledge-sharing networks. Both of these knowledge-sharing networks had similar numbers of components, similar densities and similar skewed node centrality scores. Also, these networks shared similar numbers of reach key players that could reach approximately the same proportion of other IPs in the network . These findings were surprising, as we originally thought that the longstanding statewide meeting in Iowa would have led to a much different more connected network in Iowa compared to Kentucky. Iowa does have regular APIC chapter meetings, which are similar to the structure within Kentucky. These local chapter meetings may also be a mechanism for communication in this group of professionals outside of the statewide meeting. This suggests that local APIC chapter meetings function similarly in both Iowa and Kentucky with regard to knowledge sharing. However, it appears that the only major differences in the networks are the addition of the public health professionals.
Our study has several limitations, including the possibility of missing data. As not every member of the knowledge-sharing network attended the statewide meeting, it is possible that important members of the network were not included in this analysis, thereby biasing the results. A basic assumption in knowledge sharing is that the information that is shared is correct, which may not be the case. The knowledge network may be responsible for the sharing of misinformation that may, in fact, be detrimental to HAI elimination efforts. Although generalizability to other states may be limitation, it is important to emphasize that public health professionals may play important functional roles for bolstering communication in metropolitan states as well as rural states. The ability of these professionals to focus infection prevention activities, as well as their key placement in public health departments suggests that their roles in supporting communication may apply to all types of facilities in all states.
Despite these limitations, our results demonstrate that public health officials play an important role in the communication network among IPs in Iowa. Iowa and Kentucky are similar in terms of the percentage of population living in rural regions, but in Iowa the state health department is more involved with coordinating HAI prevention. Future research efforts should be devoted to discovering what kinds of information travel across these communication networks, and understanding how the metrics we present affect real-world knowledge sharing.
In conclusion, our results suggest that the addition of public health professionals involved in the coordination of healthcare infection prevention activities may augment knowledge sharing among IPs in rural states. Given the importance of HAIs, rural states without public health involvement in infection prevention should consider the potential benefits of adding these personnel to the public health workforce.
This work was funded by a grant from the Kentucky Critical Infrastructure Protection Program, National Institute of Hometown Security (NIHS) and the United States Department of Homeland Security (DHS). The authors would like to thank Cassandra J. Wiemken, Esq. for her editorial support.
- Klevens RM, Edwards JR, Richards CL, Horan TC, Gaynes RP, Pollock DA: Estimating health care-associated infections and deaths in U.S. hospitals, 2002. Public Health Rep. 2007, 122 (2): 160-166. Epub 2007/03/16PubMedPubMed CentralGoogle Scholar
- ,: U.S. Department of Health and Human Services Office of Disease Prevention and Health Promotion. 2020, Healthy People, Washington DC, [August 17, 2011]; Available from: http://healthypeople.gov/2020/topicsobjectives2020/overview.aspx?topicid=17
- New York State Department of Health: Section 239 of the New York State Public Health Law - Course Work or Training in Infection Control Practices. [August 18, 2011]; Available from: http://www.health.state.ny.us/regulations/public_health_law/section/239/
- Pennsylvania Department of Health: Healthcare associated infection prevention. Commonwealth of Pennsylvania. [August 18, 2011]; Available from: http://www.portal.state.pa.us/portal/server.pt/community/healthcare_associated_infections/14234
- Scheckler WE, Peterson PJ: Infections and infection control among residents of eight rural Wisconsin nursing homes. Arch Intern Med. 1986, 146 (10): 1981-1984. 10.1001/archinte.1986.00360220141024. Epub 1986/10/01PubMedView ArticleGoogle Scholar
- Polgreen PM, Beekmann SE, Chen YY, Doern GV, Pfaller MA, Brueggemann AB: Epidemiology of methicillin-resistant Staphylococcus aureus and vancomycin-resistant Enterococcus in a rural state. Infect Control Hosp Epidemiol. 2006, 27 (3): 252-256. 10.1086/501537. Epub 2006/03/15PubMedView ArticleGoogle Scholar
- Scheckler WE, Peterson PJ: Nosocomial infections in 15 rural Wisconsin hospitals--results and conclusions from 6 months of comprehensive surveillance. Infect Control. 1986, 7 (8): 397-402. Epub 1986/08/01PubMedGoogle Scholar
- Scheckler WE, Peterson PJ: Nosocomial infection prevalence, risk and control in small community and rural hospitals. Infect Control. 1986, 7 (2 Suppl): 144-148. Epub 1986/02/01PubMedGoogle Scholar
- Stevenson KB, Murphy CL, Samore MH, Hannah EL, Moore JW, Barbera J: Assessing the status of infection control programs in small rural hospitals in the western United States. Am J Infect Control. 2004, 32 (5): 255-261. 10.1016/j.ajic.2003.10.016. Epub 2004/08/05PubMedView ArticleGoogle Scholar
- United States Census Bureau: census interactive population search. 2010, [August 18, 2011]; Available from: http://2010.census.gov/2010census/popmap/ipmtext.phpGoogle Scholar
- McKinney WP, Wesley GC, Sprang MV, Troutman A: Educating health professionals to respond to bioterrorism. Public health reports. 2005, 120 (Suppl 1): 42-47. Epub 2005/07/20PubMedPubMed CentralGoogle Scholar
- Wiemken T, Ramirez J, Polgreen P, Peyrani P, Carrico R: Evaluation of the knowledge-sharing social network of hospital-based infection preventionists in Kentucky. Am J Infect Control. , : -In PressGoogle Scholar
- Feltovich F, Fabrey LJ: The current practice of infection prevention as demonstrated by the practice analysis survey of the Certification Board of Infection Control and Epidemiology, Inc. Am J Infect Control. 2010, 38 (10): 784-788. 10.1016/j.ajic.2010.05.020. Epub 2010/09/25PubMedView ArticleGoogle Scholar
- Iowa Hospital Association: Profiles: documenting the social and economic importance of Iowa hospitals and health systems. 2009, , Des Moines, IAGoogle Scholar
- Knoke D, Yang S: Social network analysis. 2008, Sage Publications, Los Angeles, viii-132. 2Google Scholar
- Wiemken TL, Ramirez JA, Polgreen P, Peyrani P, Carrico RM: Evaluation of the knowledge-sharing social network of hospital-based infection preventionists in Kentucky. Am J Infect Control. 2011, : -Epub 2011/09/03Google Scholar
- Wasserman S, Faust K: Social network analysis : methods and applications. 1994, Cambridge University Press, Cambridge ; New York, xxxi-825.View ArticleGoogle Scholar
- Valente T: Social networks and health : models, methods, and applications. 2010, Oxford University Press, Oxford ; New York, xiv-277.View ArticleGoogle Scholar
- Valente T, Fujimoto K: Bridging: Locating Critical Connectors in a Network. Social Networks. 2010, 32 (3): 212-220. 10.1016/j.socnet.2010.03.003. Epub 2010/06/29PubMedPubMed CentralView ArticleGoogle Scholar
- Kleinberg JM: Authoritative sources in a hyperlinked environment. J ACM. 1999, 46 (5): 604-632. 10.1145/324133.324140.View ArticleGoogle Scholar
- Borgotti S: Identifying sets of key players in a social network. Comput Math Organ Theory. 2006, 12 (1): 21-34. 10.1007/s10588-006-7084-x.View ArticleGoogle Scholar
- Burt RS: Structural holes : the social structure of competition. 1992, Harvard University Press, Cambridge, Mass, viii-313.Google Scholar
- Burton P, Wu Y, Prybutok V: Social network position and its relationship to performance of IT professionals. Informing Science: The International Journal of an Emerging Transdiscipline. 2010, 13: 121-137.Google 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.