Dr. Oermann is Professor and Division Chair, Adult/Geriatric Health, School of Nursing, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina; Dr. Hallmark is Director of Simulation, Gordon Inman College of Health Sciences and Nursing, Belmont University, Nashville, Tennessee; Dr. Haus is Interim Director, West Penn Hospital School of Nursing, Pittsburgh, Pennsylvania; Dr. Kardong-Edgren is Assistant Professor, Washington State University, College of Nursing, Spokane, Washington; Ms. Keegan McColgan is Professor, Springfield Technical Community College, Springfield, Massachusetts; and Ms. Rogers is Professor, Carroll Community College, Westminster, Maryland.
The multisite study was coordinated through the National League for Nursing (NLN). Funding was provided by Laerdal Medical Corporation and the American Heart Association (AHA). Laerdal Medical Corporation and the AHA supplied the materials and equipment for the study at no cost to the schools of nursing that served as training sites. Representatives from the NLN, Laerdal Medical Corporation, and the AHA served on the team with the investigators who designed the study. The NLN, Laerdal Medical Corporation, and the AHA were not involved in the collection, analysis, and interpretation of data; in the writing of the manuscript; nor in the decision to submit the manuscript for publication. The investigators and site coordinators had sole responsibility for implementing the study and collecting and uploading the data. The investigators had sole responsibility for data analysis and interpretation, and for writing and submitting the manuscript. There were no restrictions by the NLN, Laerdal Medical Corporation, or the AHA on the statistical analysis or publication of the findings.
Dr. Kardong-Edgren received travel reimbursement from Laerdal Medical Corporation for speaking at a Laerdal-sponsored meeting not directly related to this work. The remaining authors have no financial or proprietary interest in the materials presented herein.
Address correspondence to Marilyn H. Oermann, PhD, RN, FAAN, ANEF, Professor and Division Chair, Adult/Geriatric Health, School of Nursing, University of North Carolina at Chapel Hill, 433 Carrington Hall, CB # 7460, Chapel Hill, NC 27599-7460; e-mail: firstname.lastname@example.org.
Decisions about educational practices should be based on evidence generated from well-designed research studies. With evidence, nurse educators can make sound decisions about teaching strategies, assessment methods, clinical practice, simulation, and other experiences for students. Nursing education has not kept pace with other nursing specialties in conducting sufficient high-quality studies to generate this evidence. Studies conducted in one setting with one small group of students are common. Many of these studies can be viewed as pilot projects. Although essential to examining the feasibility of a larger study and potentially improving its quality and logistics, the pilot may be the only study done.
Few large, multisite studies have been conducted in nursing education, and literature pertaining to implementing those studies is lacking. Barriers to multisite studies in nursing education include lack of funding; lack of expertise among nursing faculty; limited objective methods of evaluating learning and other outcomes; poorly developed and tested tools; differences in programs, learners, teachers, and other characteristics across schools of nursing that affect the ability to control variables; and issues in randomizing students to groups (Ferguson & Day, 2005; Oermann, 2009; Schmelzer et al., 2011). The purpose of this article is to describe strategies for conducting a multisite study in nursing education, using our research as an exemplar.
Overview of the Multisite Study
This study was a randomized trial to examine the effects of brief, deliberate practice on nursing students’ retention of CPR psychomotor skills. The intervention, 6 minutes of monthly practice of skills on a voice advisory manikin (VAM), was tested over 12 months in 10 schools of nursing throughout the United States. At baseline, the schools were randomly assigned to one of two types of American Heart Association Basic Life Support (BLS) courses—either HeartCode™ BLS with VAMs for learning CPR psychomotor skills or the standard instructor-led training with traditional manikins (Kardong-Edgren, Oermann, Odom-Maryon, & Ha, 2010; Oermann et al., 2010). Students (n = 616) who successfully completed one of these BLS courses were invited to participate in the yearlong study. Ten students declined to participate, and the remaining 606 were randomly assigned to the experimental or control group.
Students in the experimental group received the intervention; students in the control group had no further practice of CPR skills beyond their initial BLS course at baseline. The findings indicate that brief, self-directed practice of CPR skills, in which students were prompted by the VAM, enabled them to maintain competence and improve their performance over time (Oermann, 2011; Oermann, Kardong-Edgren, & Odom-Maryon, 2011). In contrast, students in the control (no practice) group had a rapid loss of psychomotor skills, which continued to deteriorate through the 12 months of the study.
Structure of the Multisite Study
Multisite studies in nursing education have many benefits over single-site research; the most important is that they provide a large and often diverse sample. Evaluating outcomes of interventions across multiple settings with a large sample size increases the generalizability of the findings and strengthens the evidence (Fiss et al., 2010; Flynn, 2009). However, meticulous planning and thorough preparation of those involved in a multisite study are key to the success of such a project. Initially defining the roles and responsibilities of the team members is critical to the final outcomes of the research (Weber & Cobaugh, 2008).
The study team included a principal investigator (PI); a project director; a statistician; 14 site coordinators (at 10 sites); and support staff, including an administrative assistant for the study, staff at each site, and technical support personnel. As in single-site research, the PI in a multisite study is responsible for the overall management of the study and assumes responsibility for its implementation (Flynn, 2009). This includes ensuring that the protocol is followed at each site and data are collected uniformly across sites.
Also critical to the success of this multisite study was the project director. In our study, the project director was the immediate contact person for the site coordinators. Her role was to clarify the protocol and procedures, answer questions from the coordinators and other personnel at the sites, and help problem solve and troubleshoot when unexpected issues occurred with the study.
A third member of the team was the statistician. A mistake often made by novice researchers is not consulting a statistician early in the project. Initially conferring with the statistician about the size of the sample and appropriate statistical analysis is an important step in the planning phase of the research (Weber & Cobaugh, 2008). In our study, the statistician worked closely with the PI and project director in all phases of the research, but particularly in analyzing and interpreting the data. The statistician was critical, as the analysis was complex. For each outcome measure, linear mixed models were run to examine the influence of practice (experimental or control group), initial BLS course type (HeartCode or instructor led), reassessment time (3, 6, 9, or 12 months), and all corresponding two-way interaction terms. Neither the PI nor project director was proficient in the statistical analysis required in this research.
In a multisite study, there may be a PI at each site or, as in our project, an onsite study coordinator to implement the protocol at that site. Each school participating in the study provided a site coordinator who performed the day-to-day tasks associated with the project. The coordinator was the go-to person at the site (Flynn, 2009). Some of his or her responsibilities were recruiting the participants, arranging the BLS courses at baseline, scheduling the monthly practice sessions and reassessments every 3 months, setting up and maintaining the manikins over the term of the study, troubleshooting technical difficulties with the equipment, and managing all aspects of data collection and transmission. The coordinators were key to the integrity of the data, collecting the data per protocol, preparing the electronic files for transmission to the statistician, maintaining the security of the data, and attempting to resolve issues with data files.
The roles of support staff vary according to the study design and procedures specific to it. An important resource in our study was technical support; each site had at least one technical support person who assisted with the setup, maintenance, and troubleshooting of the VAMs, data collection, and data transmission.
The site selection in a multisite study, similar to single-site research, is dependent on the research aims, study design, and other criteria specific to the study. The inclusion and exclusion criteria will influence the selection of participants and institutions. For example, in our study, participants were beginning students in prelicensure nursing programs who would be enrolled for the full length of the study (one year). To provide diversity across sites, other factors that influenced the selection of sites were the number of potential participants that would be available, demographics, and location of the institution.
To recruit sites, the team may place public announcements in professional journals or send e-mails to constituents who may be interested in participating in the research (Davidson et al., 2006). For our study, a questionnaire was sent to all members of the National League for Nursing, collecting information on the number of students in the prelicensure program available to participate, the students’ previous experience with simulation, the method of current BLS training used by the school, and other demographic information.
After the sites were selected, the coordinators were chosen by the deans or program directors. The coordinators had to be prepared to conduct the study based on the associated protocol. Training may be accomplished at a central site or may be on-site. To maintain consistency and ensure a timely launch of the study, training the coordinators at a central site at a specific time is beneficial (Nelson & Morrison-Breedy, 2008).
The initial training of our site coordinators was held at a large simulation center in Pittsburgh, Pennsylvania, for 2.5 days. The site coordinators were given detailed instructions about the research study, protocols, and schedules, and they were instructed on the use and maintenance of the VAM. Many of the nursing programs had never participated in a research study, and some site coordinators had no prior research experience, creating challenges with varying levels of understanding of the research process. The added complex study design and unfamiliar manikin technology necessitated repetition of instructions and varying methods of teaching to ensure that each site coordinator understood the process.
Site coordinators were trained with in-class PowerPoint® slides with detailed screen shots of the data collection and data uploading techniques, coupled with extensive discussion. This was followed by hands-on practice on the manikins, including the file loading and transfer protocols. All coordinators were also trained on VAM maintenance and troubleshooting software.
In addition to group training, onsite training may be needed where supplementary hands-on preparation occurs (Davidson et al., 2006). In our study, an onsite hands-on training session was provided in addition to the initial training; an educator representing the VAM vendor (Laerdal Medical Corporation) visited each site to provide more detailed education about the use of the manikins for practice and assessment of skills. The training was essential to answer questions of the site coordinators and their support staff. Fiss et al. (2010) suggested that training the researchers at each site and monitoring the work at the sites will lead to improved outcomes and assurance that the methodology is consistent across the settings.
Communication with Team
Communication in a multisite study is a challenge, but it is paramount to the success of the project. In our study, not only was effective communication necessary for the implementation of the intervention per protocol, it was important to sustain enthusiasm for the length of the study. A successful multisite study requires trust and collaboration among participating institutions and time and effort to nurture (Flynn, 2009). Team building with the site coordinators began with training at the simulation center. This initial face-to-face networking provided the foundation of rapport and trust for future communication. To maintain this relationship, further encounters of the team at nursing education conferences and working group meetings held by conference call encouraged the camaraderie and collaboration required to maintain the engagement of the site coordinators.
Traditional communication via e-mail, conference calls, fax, and postal mail were used initially to transmit procedures and answer concerns. Cell phone access to the project director provided immediate verbal responses to individual site coordinator questions. Early in the project development, it was determined that a more rapid and interactive form of communication was needed to promote collaboration and the interchange of information. To meet this need, a wiki was developed by the administrative assistant, which became a primary means of information transmission throughout the study.
A wiki is a Web-based social environment, which allows all users permission to edit the content of the site. E-mail notification of new postings and weekly updates serve to keep users informed of additions and editing changes. Not only did our wiki have value for communication, but it fostered collaboration, was a repository for forms and documents, served as a forum for sharing experiences, and saved time expended by answering repetitive questions from the larger group. Having information available in one place was critical for effective and consistent implementation in this multisite study (Kardong-Edgren et al., 2009). Frequent contact, immediate feedback, and readily accessible resources were the key communication components that fostered the success of this research project.
Recruitment and Retention of Students
The ability of researchers to attract and retain participants is a critical factor affecting the outcomes of any study. Inclusion and exclusion criteria must be clear, with a definition of who is and who is not able to participate, and in a multisite study, there must be standardization across sites in recruiting participants. After the potential population of participants is identified, there needs to be a way of encouraging an individual to make the commitment to participate, especially in a study extending for 1 year.
Many reasons exist for participants to decide to participate in a research study. Some perceive that they can derive a personal benefit from participation; others may want to contribute to learning more about a situation or condition. In our study, letters of recognition for participation were provided to the students, which could be placed in their professional portfolio. The major direct benefit for participants was that they received, at no cost to them, a 2-year BLS certification from the American Heart Association for successful completion of the course. Regardless of the incentive strategies used, there needs to be a clear understanding of what the project entails and the benefits and risks involved with it.
No single model exists for study recruitment and retention, as each research project has its own challenges. Site coordinators in our study used various means of communication, including e-mails and colorful flyers, to attract students to join the study. One site coordinator recruited during a class on professionalism in nursing. The request for participation coincided with what the students were learning at the time. The importance of research in nursing was presented, and students were informed of the opportunity to learn about the research process by actually participating in a study. Research designs, standardization of data collection procedures, and informed consent are a few of the concepts that students could learn through involvement in our study. Face-to-face discussions with potential participants helped with informed consent by clarifying issues and concerns they had at the time. This method also contributed toward developing personal relationships between the site coordinators and participants, which promoted a commitment to study completion.
With a longitudinal study, retention of participants can be challenging. It is important to retain as many participants as possible because a large sample from multiple sites enhances external validity and allows for greater generalizability than a single site study (Gold & Dewa, 2005). Statistical power is dependent on participant retention, and attrition can have serious effects on the outcomes of the study. Researchers go to great lengths to recruit participants and need to expend as much, if not more, effort in retaining them.
The site coordinator had a critical role in participant retention. Positive reinforcement and comments about the importance of students’ continued participation during each encounter appeared to motivate participants to remain in the study. Site personnel can combat wavering interest of participants by maintaining and exhibiting excitement for the research (Berger, Neumark, & Chamberlain, 2007).
Other strategies that were successful in retaining students in our study included developing ways to make participation as easy as possible, with the least amount of disruption to students’ schedules. For example, blocks of time were posted during which participants could sign up for a practice session or reassessment; the time slots coincided with student breaks from class or were immediately before or after class to avoid a special trip to the school. Participants were sent reminders via e-mails or notes placed in their mailboxes about signing up for the upcoming sessions. Communication was critical in keeping students coming back. One school offered community service hours for students’ participation in the study.
Figure 1 presents the flow and number of participants retained in each phase of the study. One hundred eleven students dropped out before the final BLS testing. Some students did not continue in the study due to a lack of interest, whereas others were lost because of attrition from the nursing program. We believe our personal contact and individualized attention to the needs of participants accounted for our high retention rate (N = 495, 81.7%). Evidence suggests that the relationship of the study personnel with the participants may be the single most important factor in recruitment and retention (Penckofer, Byrn, Mumby, & Ferrans, 2011).
Figure 1. Flow and number of participants throughout the study. From “Effects of monthly practice on nursing students’ CPR psychomotor skill performance,” by M.H. Oermann, S. E. Kardong-Edgren, and T. Odom-Maryon, 2011, Resuscitation, 82, p. 448. Copyright 2011 by Elsevier. Reprinted with permission. Note. BLS = basic life support.
Technical Aspects of the Study
Because our intervention involved the use of technology, site coordinators were responsible for ensuring that the laboratories used for implementing the study had adequate floor space, electrical outlets, Internet access, and a secure area for storage of equipment. The VAMs with connecting laptops required access to electrical outlets and the Internet (Figure 2). Not every school had dedicated space assigned for study implementation. Sites that had dedicated space had more control over scheduling of CPR practice and testing sessions and could ensure consistent electrical and Internet access. Sites were assigned a specific number of VAMs based on their number of participants, ranging from one adult and one infant VAM for one school to five adult and five infant VAMs in other schools. The number of VAMs affected the size of the secure areas in the laboratory in which students practiced and underwent testing. The technical support staff at each of the sites provided the necessary support for both the initial setup of the VAMs with connecting laptops and throughout the duration of the study. Providing a secure area to house the VAMs and laptops also was challenging for some sites. A few sites used locked boxes in areas adjacent to the testing areas, which safely stored both VAMs and laptops, allowing easy access for site coordinators.
Figure 2. The voice advisory manikin, with connection to electrical outlets and laptop computer with Internet.
Data Collection, Management, and Quality
After students consented to participate in the yearlong study, they were randomly assigned to the intervention or control group and a reassessment month. Site coordinators provided chronologically numbered individual envelopes as the students signed the consents. The students opened the envelopes to reveal the group to which they had been randomly assigned. This study number had matching envelopes with preplanned study activities generated by the statistician for every month of the study. Students returned each month for the next envelope in their sequence to discover their next steps in the study: practice or test out for the experimental group, or return next month or test out for the control group. Site coordinators were masked to the history associated with each study number. Every study number had 12 envelopes stuffed with paper to maintain the masking of the site coordinators and students (i.e., if numbers and envelopes stopped at a certain point, both coordinators and participants would know they were finished with the study at that point in time).
In our study, the data were collected electronically via the VAM using software that measured the quality of the students’ performance of compressions, ventilations, and single-rescuer CPR on the manikin. Site coordinators made individual online file folders for recording these data for each participant at baseline and at reassessment. Coordinators had to enter the participant’s six-digit code into the program each time a data file was collected; this included three individual files (compression, ventilation, and single rescuer CPR) at baseline and at reassessments for every student who participated at their site. This proved challenging, as one digit out of place necessitated reviewing the individual files, identifying the error in any number of the Excel® spreadsheet columns, correcting the error, and renaming the file in another data bank. These tasks were shared by the statistician and project director. If a file error was identified early, the student would be retested. Another challenge with the data collection was that site coordinators often worked alone and had to find time to upload the multiple student files.
Due to the complexity of the study design, some protocol errors were made. For example, in month 3, one site coordinator tested students in the 6-month and 9-month groups, as well as the scheduled 3-month test out group. A quick telephone call refreshed the site coordinator’s understanding, which cleared up misconceptions that had existed about the protocol. We learned from this experience that it was useful to ask each site coordinator to explain the complete study design and procedure aloud to confirm their understanding of the protocol.
The statistician reviewed incoming data to ensure quality control and minimize errors. The mislabeling of individual data files was common and time consuming to correct. For this reason, an additional customized data management program was commissioned from an outside vendor to assist with file management and data error identification. Use of the wiki, interspersed with phone calls, assisted the study team to stay in communication during the data collection phase of the study.
Ethical Considerations and Budgeting
All human subject research studies, whether single site or multisite, require institutional review board (IRB) approval. In a multisite study, all sites do not need their own IRB to approve the study if deferral agreements can be negotiated. For our study, if we did not pursue such an agreement, IRB approval would have been needed from each of the 10 institutions. As Gold and Dewa (2005) suggested, pursuing 10 approvals would take excessive time, effort, and resources and could lead to delays and potential inconsistencies in how participants were treated across sites. One of the challenges related to the IRB process that Flynn (2009) encountered in her study of medication errors in 14 acute care hospitals was that the IRB in each of the study hospitals had a different process for review and different forms to use.
The alternative approach was to apply for IRB approval as a multisite study at the PI’s institution, serving as the lead coordinating center responsible for the overall study conduct, with the other sites deferring to the coordinating site. This approach expedited the IRB review process and avoided inconsistencies across sites. The PI completed an addendum that applied to multisite studies, with each of the sites then providing documentation that the study would be conducted at their location in compliance with the regulations and ethical principles governing human subjects research. The site coordinators ensured the study was implemented in their institution as approved. During the IRB process, the site coordinators worked closely with the PI and the IRB committees in their own settings to facilitate approval as multisite study.
Currently, few federal agencies fund educational research. Educational grants from professional organizations, companies that develop manikins and other products, and some foundations may not allow indirect costs (expenses of implementing the grant other than salaries, materials, travel, and other direct costs) to be calculated into funding requests. Some educational grants do not provide researcher salary support nor do they cover other costs of implementing the study. In planning the budget, it is critical to identify all costs for the study across sites. In some instances, nursing faculty members may negotiate reduced teaching time for a semester or year and obtain internal grants to cover any additional costs. As the use of technology increases in nursing education, more large scale studies to test those innovations need to be conducted and adequately funded.
Regardless of the type of study, dissemination is critical to communicate the findings to others and is the last phase of a research project. With multisite projects, decisions about authorship and manuscript submission should be made at the outset to avoid conflict among members of the team. With our project, the PI identified topics for manuscripts and, for selected papers, invited site coordinators to participate with the writing of them. For most coordinators, this was their first experience in writing for publication.
Because the study was a randomized trial, the manuscript on the main findings was prepared using the Consolidated Standards of Reporting Trials (CONSORT) Statement (CONSORT, 2011). The CONSORT Statement provides guidelines for writing the title, abstract, introduction, methods, results, and discussion sections of the manuscript. As recommended, we included a flow diagram that depicted how participants progressed through the trial in the enrollment, allocation to the intervention and control groups, and follow-up reassessment phases (Figure 1). The CONSORT Web site ( http://www.consort-statement.org/consort-statement/) provides a downloadable template for preparing the flow diagram for a manuscript, which we used. Other manuscripts were prepared to reach particular audiences, for example, BLS educators in hospitals. One of the funding agencies released short summaries of the study findings, written for the public, providing another opportunity for dissemination.
Evidence for educational decisions in nursing depends on findings generated from high-quality and relevant research studies. Multisite studies provide large samples and improve generalizability, but they create different challenges for nursing education researchers than with projects implemented in one setting. However, multisite studies provide an opportunity to test interventions with multiple groups of students across schools of nursing. This article describes strategies used for conducting a multisite study in nursing education, which readers may find valuable if they are planning a research project to be implemented in multiple sites.
- Berger, A.M., Neumark, D.E. & Chamberlain, J. (2007). Enhancing recruitment and retention in randomized clinical trials of cancer symptom management. Oncology Nursing Forum, 34(2), E17–E22. doi:10.1188/07.ONF.E17-E22 [CrossRef]
- CONSORT. (2011, April22). The CONSORT statement. Retrieved from http://www.consort-statement.org/consort-statement/
- Davidson, R.M., McNeer, J.F., Logan, L., Higginbotham, M.B., Anderson, J., Blackshear, J. & Wagner, G.S., . . . (2006). A cooperative network of trained sites for the conduct of a complex clinical trial: A new concept in multicenter clinical research. American Heart Journal, 151, 451–456. doi:10.1016/j.ahj.2005.04.013 [CrossRef] doi:10.1016/j.ahj.2005.04.013 [CrossRef]
- Ferguson, L. & Day, R.A. (2005). Evidence-based nursing education: Myth or reality?Journal of Nursing Education, 44, 107–115.
- Fiss, A.L., McCoy, S.W., Bartlett, D., Chiarello, L.A., Palisano, R.J., Stoskopf, B. & Wood, A., . . . (2010). Sharing of lessons learned from multi-site research. Pediatric Physical Therapy, 22(4), 408–416. doi:10.1097/PEP.0b013e3181faeb11 [CrossRef] doi:10.1097/PEP.0b013e3181faeb11 [CrossRef]
- Flynn, L. (2009). The benefits and challenges of multisite studies: Lessons learned. AACN Advanced Critical Care, 20(4), 388–391. doi:10.1097/NCI.0b013e3181ac228a [CrossRef] doi:10.1097/NCI.0b013e3181ac228a [CrossRef]
- Gold, J.L. & Dewa, C.S. (2005). Institutional review boards and multisite studies in health services research: Is there a better way?Health Services Research, 40, 291–308. doi:10.1111/j.1475-6773.2005.00354.x [CrossRef] doi:10.1111/j.1475-6773.2005.00354.x [CrossRef]
- Kardong-Edgren, S., Oermann, M., Ha, Y., Tennant, M., Snelson, C., Hallmark, E. & Hurd, D., . . . (2009). Using a wiki in nursing education and research. International Journal of Nursing Education Scholarship, 6(1), Article 6. Retrieved from http://www.bepress.com/ijnes/vol6/iss1/art6/ doi:10.2202/1548-923X.1787 [CrossRef] doi:10.2202/1548-923X.1787 [CrossRef]
- Kardong-Edgren, S., Oermann, M.H., Odom-Maryon, T. & Ha, Y. (2010). Comparison of two instructional modalities for nursing student CPR skill acquisition. Resuscitation, 81, 1019–1024. doi:10.1016/j.resuscitation.2010.04.022 [CrossRef]
- Nelson, L.E. & Morrison-Breedy, D. (2008). Research team training: Moving beyond the job descriptions. Applied Nursing Research, 21(3), 159–164. doi:10.1016/j.apnr.2006.09.001 [CrossRef] doi:10.1016/j.apnr.2006.09.001 [CrossRef]
- Oermann, M.H. (2009). Evidence-based programs and teaching/evaluation methods: Needed to achieve excellence in nursing education. In Adams, M. & Valiga, T. (Eds.), Achieving excellence in nursing education (pp. 63–76). New York, NY: National League for Nursing.
- Oermann, M.H. (2011). Toward evidence-based nursing education: Deliberate practice and motor skill learning. Journal of Nursing Education, 50, 63–64. doi:10.3928/01484834-20110120-01 [CrossRef] doi:10.3928/01484834-20110120-01 [CrossRef]
- Oermann, M.H., Kardong-Edgren, S., Odom-Maryon, T., Ha, Y., McColgan, J.K., Hurd, D. & Dowdy, S.W., . . . (2010). HeartCode™ BLS with voice assisted manikin for teaching nursing students: Results of a preliminary study. Nursing Education Perspectives, 31, 303–308.
- Oermann, M.H., Kardong-Edgren, S.E. & Odom-Maryon, T. (2011). Effects of monthly practice on nursing students’ CPR psychomotor skill performance. Resuscitation, 82, 447–453. doi:10.1016/j.resuscitation.2010.11.022 [CrossRef] doi:10.1016/j.resuscitation.2010.11.022 [CrossRef]
- Penckofer, S., Byrn, M., Mumby, P. & Ferrans, C.E. (2011). Improving subject recruitment, retention, and participation in research through Peplau’s theory of interpersonal relations. Nursing Science Quarterly, 24, 146–151. doi:10.1177/0894318411399454 [CrossRef] doi:10.1177/0894318411399454 [CrossRef]
- Schmelzer, M., Daniels, G., Dudley-Brown, S., Foote, K., Hayes, A. & Nardelli, N. (2011). Multisite research: A spirit of cooperation. Gastroenterology Nursing, 34, 327–330. doi:10.1097/SGA.0b013e318227cbd9 [CrossRef] doi:10.1097/SGA.0b013e318227cbd9 [CrossRef]
- Weber, R.J. & Cobaugh, D.J. (2008). Developing and executing an effective research plan. American Journal of Health-System Pharmacy, 65, 2058–2065. doi:10.2146/ajhp070197 [CrossRef] doi:10.2146/ajhp070197 [CrossRef]