NLINKS - Language In Nursing Knowledge Systems

Current Study

The subjects of diagnoses being reviewed in the first study for the annotated bibliographic database are sleep deprivation and disturbed sleep pattern (NANDA, 2001-2002).

Sleep is chosen as the subject of the first review because:

There are NANDA diagnoses on sleep (i.e., disturbed sleep pattern and sleep deprivation).
Sleep is a human response that crosses all nursing specialities.
There is a relatively long history on the contribution of nursing to the sleep literaure. Electronic databases that include sleep reference nursing articles that date back to the 60s.

No only NANDA diagnoses and related interventions and outcomes will be reviewed. In fact one of the advantages of this methodology, once refined, is that it will allow nurses to systematically review and code articles dealing with nursing diagnoses not as yet classified. Thus, the method will directly benefit language development in nursing.

Award Amount/Source of Support
$4,300 Beaumont Faculty Development Fund Grant,
Saint Louis University, from July 1, 2001 through
June 30, 2002.

The purpose of this grant is to establish an online, peer-reviewed annotated bibliography to serve as a first step in the synthesis of nursing research if the field of nursing diagnoses and nursing language development. Nursing diagnoses refer to the clinical judgments that nurses make. Nursing language refers to standardized terms used to represent nursing diagnoses, interventions, and outcomes. To accomplish this objective, an interuniversity team of co-investigators developed a proposal, funded by the Beaumont Faculty Development Fund of Saint Louis University.

Study process
In the study's initial phase, the investigators will solicit 150 bibliographic reviews through this site, the NLINKS Research Center. After studying the reviews submitted and the articles annotated, the team will develop criteria against which each review is to be evaluated. Once the evaluation criteria are established, the study's second phase begins with the investigators calling for a new submission of annotated references. At the end of this phase, accepted annotated bibliographic reviews and their evaluations will be published online. This concludes the second and last phase of the study. At this point, a mechanism will have been established for the continued solicitation, evaluation, and publishing of annotated bibliographic references. These peer-reviewed annotations will assist users in selecting references for their own reviews, serve as a step in the development of an online method of synthesizing research to use in the development of evidenced-based practice recommendations, and stimulate study and research in the field.

Future funding for the database
Future funding for the annotated bibliographic database will be sought from a variety of nursing, library, and informatics resources. Of these first and foremost is the National Library of Medicine.

Investigators
Mary Ann Lavin, DSc, RN, BC, ANP, FAAN, Associate Professor and
Principle Investigator, Saint Louis University

Judith Carlson, MSN, RN, Associate Professor, Saint Louis University

June Clark, PhD, RN, FHU, FRCN, Professor,
University of Wales-Swansea

Dotti James, PhD, RN, Assistant Professor, Saint Louis University

Mary Krieger, MLIS, RN, Health Sciences Center Library,
Saint Louis University

Meridean Maas, PhD, RN, FAAN, Professor, University of Iowa

Patricia McNary, MALS, RN, Health Sciences Center Library,
Saint Louis University

Geralyn Meyer, PhD, RN, Saint Louis University

Anne G. Perry, EdD, RN, Professor, Saint Louis University

Stephanie Richardson, PhD, RN, Assistant Professor,
University of Utah

Sarah Rodseth, SN, University of Utah

Search strategies
Topic of review:
Disturbed sleep pattern and sleep deprivation.

Search Strategies:
Two search strategies were developed. One for PubMed and one for CINAHL. The latter references were converted by Reference Manager for Windows and the resulting list named Bibiography for Disturbed Sleep Pattern and Sleep Deprevation.

Availability of Search Strategies:
A copy of these strategies may be obtained by email from the two Health Science Centers libraries at Saint Louis University.

Notes on strategies:
Pub Med - notify NLINKS by email of the article you selected. Bibliography for Disturbed Sleep Pattern and Sleep Deprevation covers years 1982 to the present. A mechanism is being developed to identify articles that have been selected.

Preventing duplication of effort.
It is conceivable that two reviewers or two review groups may select the same article(s). During this pilot phase, both reviews will be submitted to the investigators for their decision. However; other steps are being taken to prevent duplication.

Progress Reports

Progress Report of August, 2004
The following databases and a pop-up box user’s guide for each database are now complete:

The Nursing Diagnosis Database
Nursing Diagnosis and Primary Data Database
Nursing Sensitive Patient Outcomes Database
The Primary Data Database
Nursing Sensitive Patient Outcomes and Primary Data Database

Your feedback is appreciated. Contact: webmaster@nlinks.org

Progress Report of March, 2003
The nursing sensitive Patient Outcomes Database is complete. Just click on this hyperlink and it will appear via PubMed. Once the PubMed search appears, you may limit it in any way you like; or, you can add it to any other search term or search string. For example, if you would like to obtain a search on patient outcomes associated with therapeutic touch interventions, just add to the patient outcomes search string the following. Be sure to conclude with the Boolean AND:

(therapeutic touch OR therapeutic touch interventions) AND

Progress Report of November, 2002
Much has occurred since the NLINKS Progress Report in June, 2002. A summary follows.

Publications
Lavin, MA, Meyer, G., Krieger, M., McNary, P, Carlson, J., Perry, A., James, D., Cvitan, T. (2002). Viewpoint: essential differences between evidence-based nursing and evidence based medicine. International Journal of Nursing Terminologies and Classifications (July-September) Vol 13(3):101-106.

Scheduled presentations on one or more aspects of the EBN Matrix.
Nov. 14-17, 2002
Policy Forum of the Nursing Organization Alliance (NOA) Inaugural Meeting. "NLINKS and its Evolving Nursing Diagnosis and Patient Outcomes Databases: Global Research Implications." Indianapolis, Indiana. Presenter: M.A. Lavin, ScD, RN, FAAN

March 20-22, 2003
"Patient outcome filter and database development, Fourth European Conference of ACENDIO: Making nursing visible, Paris, France. Primary presenter: M.A. Lavin, ScD, RN, FAAN

June 5-7, 2003
"Evidence-base of nursing classification: Analysis and recommendations for the future" and "Describing nursing effectiveness through standardized nursing languages and computerized clinical data," Baskent University, Ankara, Turkey. Primary presenter: M.A. Lavin, ScD, RN, FAAN

June 14, 2003
"Nursing diagnoses in the 21st century and their evidence base," Japan Society of Nursing Diagnoses, Fukuoka International Conference Center, Fukuoka, Japan. Primary presenter: M.A. Lavin, ScD, RN, FAAN

On-going research projects
Development of a patient outcomes database filter:
Sensitivity and specificity analyses, Beaumont Grant Faculty Development Team, Saint Louis University, USA

Application of the patient outcomes database to the EBN-2 Scale of the EBN Matrix, Geralyn A. Meyer, PhD, Saint Louis University, USA

Evaluation of EBN-Matrix databases:
Interrater reliabilities, Anne G. Perry, EdD, Saint Louis University, USA

Initial testing of the primary data and nursing diagnosis database filters of the EBN Matrix, Cordie Reese, EdD, Saint Louis University, USA

Research ideas in development:
Review of a PubMed nursing diagnosis search as applied to the EBN Matrix and subcategory development, Almuth Berg and Steffen Fleischer, Master's Degree Students, Martin Luther University Halle-Wittenberg, Institute for Nursing and Health Sciences, Halle, Germany

*If you would like to contribute to the development of EBN-Matrix filters and resultant databases, join NLINKS and contact us.

Progress Report of June, 2002 This progress report is stored as a pdf document, and can be found here.

Progress Report of May, 2002
Reflecting upon the essential differences between EBN and EBM, the Evidence Based Nursing (EBN) Research Team decided to define its own EBN levels of evidence in a manner that departed from EBM. These levels are not based on studies per se but on data. They are: primary, secondary, and tertiary data tiers.

Primary data refers to data collected at the point of patient contact, whether the data is quantitative or qualitative.
Secondary data refers to analyses made on the basis of secondary datasets, e.g, the Minimum Nursing Dataset, or secondary analysis of patient data presented in the literature, e.g., decision analysis, cost-effectiveness studies, systematic reviews, meta-analyses.
Tertiary data refers to patient or client information derived from expert opinion articles, expert opinion surveys, letters to the editor, case vignettes but not actual cases.

Only the primary data filter was developed in this preliminary work. The following EBN Matrix emerged, where x represents the diagnosis filter, y represents the primary data filter, and z represents the results the application of the x and y filters to the search strategy. These filters were applied to the focused sleep search database of MEDLINE articles obtained through PubMed search engine of the National Library of Medicine, Washington, DC.

Evidence Level	Diagnosis	Related Factors	Diagnostic Tests	Interventions	Outcomes	Total
Primary Data	z					y(1)
Secondary Data
Tertiary Data
Total	x(2)

(1) The actual y filter (primary data filter as of March 28, 2002) follows: (randomized controlled trial* OR clinical trials OR clinical trial* OR single blind* OR double blind* OR triple blind* OR unblind* OR prospective study OR prospective studies OR case study OR case studies OR case control study OR case control studies OR case series OR "quasi experimental" OR pilot study OR pilot studies OR "experimental study" OR "experimental studies" OR crossover design OR "qualitative study" OR ethnograph* OR phenomenolog* OR ethnonursing OR grounded theory OR controlled clinical trial OR controlled clinical trials OR bias OR reproducibility of results OR research) NOT (review OR review[pt]).
(2) The actual x (diagnosis filter, as of March 28, 2002) filter follows: (diagnosis OR defining characteristic* OR signs and symptoms OR diagnostic concepts OR "nursing assessment" OR "diagnostic term" OR "assessment measure" OR "assessment measures" OR differential diagnosis OR "clinical judgment*" OR "clinical decision making"OR "human response patterns" OR diagnostic errors OR "functional health patterns" OR "clinical assessment tools" OR nursing diagnosis OR "diagnostic terms").

When applied to the 805 focused sleep search conducted on April 29, 2002, the y filter retrieved 260 articles sleep articles derived from primary data sources. When applied to the 805 articles obtained through the focused sleep search on April 29, 2002, the x filter retrieved 547 articles dealing with the diagnosis of sleep. The end product z, as obtained on April 29, 2002, consisted of 183 citations obtained after the diagnosis and primary data filters were applied to the 805 focused sleep articles.

Progress Report of April, 2002
NLINKS and Nursing's Agenda for the Future. A press release of April 4, 2002 states, "Nursing's Agenda for the Future Calls for Broad Support from Other Stakeholders." To read the full press release, access it here. To learn how NLINKS fits in with Nursing's Agenda for the Future read on.

NANDA is one of the stakeholders mentioned above, and NLINKS provides the infrastructure to accomplish the tasks assigned to NANDA. Specifically, NLINKS will establish a central database/clearing house/resource center for sharing best practices (Delivery Systems/Nursing Models Domain) and for sharing workforce and patient outcomes databases (Legislation/Regulation/Policy Domain). This work of NANDA was approved by the NANDA Board in an electronic vote in December, 2001 and was confirmed at the February 2002 Board Meeting in Washington, DC.

As viewers may realize, evidence based literature filtered within today's most commonly used databases could be more sensitive to nursing research. The NLINKS team has developed three nursing sensitive evidence based tiers or levels call primary, secondary, and tertiary patient data that can be cross indexed with contribution to nursing categories. This work has resulted in a "primary patient data" filter and "nursing diagnosis" filter through which articles derived from a search strategy may be passed. These findings will be presented at the April 9 - 13 NNN Conference in Chicago.

In addition, the team has begun working on secondary and tertiary patient data filters. Once these are developed, then the team will create a "nursing outcome" filter. At that point, databases commissioned by Nursing's Agenda for the Future can be added. These databases are three:

Patient outcomes
Workforce outcomes
Best practices

Finally, the team will be create strength of recommendation levels for nursing's evidence base. In the meantime, guidelines and best practices will be collected from the various specialty nursing organizations and categorized in such a way that they are readily accessible

Some may disagree with this approach. They may say that we in the nursing profession do not need to recreate the wheel. In other words, they say that medicine has already developed evidence based criteria. Nursing just needs to apply what is already there. However, the filters that are already there kick out much of nursing research or map evidence based nursing incorrectly. The reason is not that nursing research as it stands is invalid nor impossible to map. The reason is that the filters used were not developed with the intent of mining nursing research literature fully. Nor were the filters developed in a manner that takes into consideration essential differences between medical and nursing research.

Although these differences are subtle, they are real. Three differences are research focus, methods, and subject matter. Medicine focuses primarily on treatment effects or outcomes while nursing focuses primarily on patient outcomes, whether those outcomes be the presence or absence of tracheobronchial aspiration or outcomes that center on reports of a client's experience of health, hospitalization, birthing, etc. The methods that represent the highest level of evidence in medicine are limited to meta-analyses, double blind randomized control trials, randomized control trials, controlled trials, cohort studies, and case series. Methods employed at the highest levels of nursing research are quantitative and qualitative studies along with cohort, case series, and individual case studies. The subject matter of medical research is disease; the subject matter of nursing research is human response to health or illness. Because of these real differences, filters different than medicine's are needed to access scientific evidence in nursing

Again, there still may be some who still say that medicine has already created scientifically sound evidence based filters that nursing need only apply. The NLINKS research team thinks otherwise and we are not surprised that the evidence based filters we developed for nursing are similar but different from medicine's. To paraphrase Florence Nightingale in her Notes on Nursing (p. 101 - 102): We need not do everything that medicine does without regard to whether it is the best that nursing can do.

Progress Report of February, 2002
Since September, 2001, the investigators at Saint Louis University School of Nursing tested the reliability of the Evidence Based and Contribution to Nursing Language Scales. The Evidence Based Scale contains 15 choices and the Nursing Contribution Scale 17 items. Five investigators read and scaled two articles (Davignon & Bruno, 1982; Gould, 1983). Except for one outlier score for each article, agreement among investigators was more than 80% on each item scored on the Contribution to Nursing Language Scale in the Davignon article and 100% on each item scored in the Gould article. In other words, there was a high level of agreement among the investigators as to whether or not the article contributed to the development of diagnostic, intervention, or outcome language. Reliability of the Evidence Based Scale was considerably less impressive. With each article, agreement among the five investigators was equal to or less than 40% on each item scored. As a result, a team of two investigators re-examined the evidence based medical literature to explore why the scores where unimpressive and to study how they might improve the evidence-based scale. They:

reviewed the glossary of terms developed by the Evidence Based Medicine Center at the University Network-Mount Sinai Hospital in Toronto, Canada (http://www.cebm.utoronto.ca/glossary/); and,
compared the levels of evidence as described by the Centre for Evidence-Based Medicine at Oxford University (http://cebm.jr2.ox.ac.uk/docs/levels.html) and the University of Illinois-Chicago (http://www.uic.edu/depts/lib/lhsp/resources/levels.shtml).
developed a three-tiered evidence-based nursing scale, where the tiers represent primary, secondary, and tertiary data.

A grid is now under construction. Other tasks...

develop a glossary of terms for each of the items used in the grid
identify filters to be used to search the literature within each cell
test the end-product using the same PubMed search strategy on sleep disturbance/deprivation previously discussed.

Progress Report of September, 2001
In a meeting of investigators at Saint Louis University on August 23, 2001, the 17 articles that had been reviewed were discussed. The focus of all articles was intended to be disturbed sleep patterns and sleep disturbances. The investigators noted that each reviewer used a slightly different search strategies. Therefore, the librarian-investigators assumed responsibility for developing a search strategy to be used consistently throughout the remainder of this study.

On September 14, 2001, the Saint Louis University investigators again met. Two search strategies were presented. The first is a CINAHL search strategy, covering the years 1982 through 2001. This search can be accessed automatically through OVID at Saint Louis University. The second search strategy is a MEDLINE search of articles for the years 1966 to 1982, accessed through NLINKS by means of a hyperlink with PubMed at the National Library of Medicine at the U.S. National Institutes of Health.

Those investigators not located at Saint Louis University will receive a list of articles obtained by using this search strategy for the years they are assigned. They will, in turn, select five articles to be reviewed from the list, complete the ECNL form for each article, and mail or fax the completed form and one copy of the article to M. A. Lavin, ScD, RN, FAAN, Saint Louis University School of Nursing, 3525 Caroline Mall, Saint Louis, Missouri 63104, USA (fax: 314-577-8949; phone: 314-577-8961).

Legal Information
For more information, please call (314)-577-8961 or email info@nlinks.org
Web site specific questions? Contact webmaster@nlinks.org