Minimum Data Sets for Nursing
- What is a minimum data set?
- US Nursing Minimum Data Set
- Nursing Management Minimum Data Set
- International Nursing Minimum Data Set
- Canadian Nursing Minimum Data Set
- Additional Resources for NursingMinimum Data Sets
What is a minimum data set?
A minimum data set is the designation of categories with agreed upon description of exactly what type of data the category will contain that will provide the minimum data needed for a given purpose. Because minimum data sets serve different aims, there are numerous minimum data sets in healthcare such as the Outcome and Assessment Information Set (OASIS) for Home Healthcare Agencies and the Minimum Data Set (MDS) for reporting the assessment of residents in Medicare or Medicaid certified nursing facilities. In addition, many states have required minimum data sets in healthcare.
Most of the minimum data sets also include the terminology to be used to report the data for each category, although this is not a requirement. A minimum data set must meet the following criteria:
- "The data must be useful to potential users, for example, legislators, health care professionals, administrators, or regulatory bodies.
- The items in the set must be readily collectable with reasonable accuracy.
- The items should not duplicate other available data.
- Confidentiality must not be violated."(1 p 207)
In addition, Goossen et al (1998) listed five aspects that are also necessary in an NMDS
- Pertinent data items need to be identified.(2)
- Each item must be accurately defined, not only what it is, but what it is not.
- A list of possible values for each item must be determined. These can be lists of agreed terminology.
- It must be possible to document in the patient record with the appropriate terminology each variable (item).
- It must be possible to aggregate the data from individual records and used beyond the care of one patient.
References
1 Thede LQ. Informatics and Nursing: Opportunities & Challenges. 2nd ed. Philadelphia: Lippincott Williams & Wilkins; 2003.
2 Goossen WTF, Epping PJMM, Feuth T, Dassem TWN, Hassman A, VacDenHeuvel WJA. A comparison of nursing minimal data sets. Journal of the American Medical Informatics Association. 1998;5(2):152-63. Open access at http://www.ncbi.nlm.nih.gov/pmc/articles/PMC61286/
United States Nursing Minimum Data Set
United States Nursing Minimum Data Set
The United States Nursing Minimum Data Set (NMDS) was birthed in the atmosphere created in 1965 by the Social Security Act Amendments that initiated the Medicare and Medicaid programs(1). Efforts to identify national health data standards and guidelines led to the development in 1969 of the uniform minimum health data set (UMHDS). (2) Following this, the National Committee on Vital and Health Statistics (NCVHS) guided the development of three patient-focused health data sets: the Uniform Hospital Data Discharge Set, the Long-Term Healthcare Minimum Data Set, and the Uniform Ambulatory Medical Care Minimum Data Set, none of which contained nursing data.
This lack was discussed at the Nursing Information Systems conference held in 1977 at the University of Illinois College of Nursing(2). They agreed that there was a need to provide nursing data for a healthcare minimum data set. Nurses in attendance were also encouraged to submit proposals to develop nursing information systems.(3)
Eventually, in 1983, spearheaded by Harriet Werley, a professor at the University of Wisconsin - Milwaukee (UW-MI), with support from her Dean, Norma Lang, a planning committee for a conference to create a Nursing Minimum Data Set (NMDS) was created.(1) There were two main groups interested: nursing administrators who needed to account for nursing care delivery, and faculty who were beginning to research the clinical care nurses provided. After many meetings and discussions the group arrived at a consensus that the goal of an NMDS was professional control of practice. They believed that the way to control practice was to control information. An appealing prospect to administrators for creating the NMDS was the thought of being able to realize nursing services as revenue generating, thus forcing hospitals to negotiate with nursing.(1) Searches for funding for the conference, after many turn downs, eventually resulted in a positive response from the Hospital Corporation of America Foundation and IBM, and support from the University of Wisconsin School of Nursing. It was held in Milwaukee in May 1985.
Attending this conference were sixty-five invited individuals representing experts in nursing, information systems, and health records.(1) There were thirty commissioned papers that were used as jumping off points for the conference discussion. The topics included data set integration, regulatory requirements, and maintenance of the NMDS. Using these papers as a background the group developed task forces around six areas: demographics, assessment, diagnosis, intervention, outcomes, and acuity/intensity.
After a summer spent reviewing papers resulting from the conference, and heated exchanges between Professor Werley and Dean Lang, a post-conference was held in Chicago in September, 1985.(1) The issues unresolved by the task force leaders created the agenda for this conference. One of Professor Werley’s main themes was “...that the data set was not designed to support clinical decision making, but rather was to represent clinical practice after it was delivered.” (1) p 144 in the 19th volume of Nursing History Review and 18 in pdf] There was much discussion about the data to use for the nursing diagnosis. The North American Nursing Diagnosis Association (NANDA) at the time was pushing the American Nurses Association to support NANDA for all nursing diagnoses and wanted NANDA declared the official data for use in collecting nursing diagnosis. However, others who believed that there were other settings for healthcare than hospitals wanted the Omaha System considered. (In 1985 the ANA had not yet started to recognized nursing terminologies, and the only ones available were NANDA and the Omaha system.)
Due to the inability to resolve this issue, there were no decisions made about the data for the nursing diagnosis category. Additionally, no decisions were made about the data to use for interventions or outcomes, which is why the data set only specifies the type of data to be collected along with definitions of each type, not the terminology to use to collect the data. The US NMDS remains unchanged to this day and consists of 16 elements. It was derived from the general concept of a Uniform Minimum Health Data Set.(3) The end result was 16 elements in three categories as seen in figure 1.
Nursing Care Elements |
Patient Demographic Elements |
Service Elements |
*Nursing Diagnosis |
*Unique Number of Principal *Registered Nurse Provider Date of Birth |
Unique facility or agency number elements |
Figure 1 Elements of the NMDS (4)
The elements listed marked with an asterisk (*) are the only ones that are original in this data set. The others are in the Universal Hospital Discharge Data Set (UHDDS).
With the exception of intensity of nursing care, the nursing care elements in the NMDS are the basis for the three categories in standardized nursing terminologies: problem or diagnosis, intervention, and outcome. As described in the NMDS data collection manual, intensity was to be determined by the amount of nursing care administered to the patient and the ratio of various types of nursing personnel involved in the patient’s care.(4) However this was never standardized. Attempts have been made by many agencies and vendors to standardize nursing intensity for both staffing purposes and billing.(5) Interestingly, several authors have found that adding nursing diagnosis to DRGs provides a more accurate picture of nursing intensity than the DRGs alone. (6),(7)
The duplication of items in the NMDS was of concern to many, and became a major hurdle in Werley’s fight to have the government require use of the data set. Using her many accomplishments, and government connections including Faye Abdellah, who at that time was deputy surgeon general and chief nurse officer of the U.S. Public Health Service, a hearing before the Department of Health and Human Services was arranged. (1) The committee commented on the NMDS being specific to only one profession instead of the practice setting, the duplication of data, and the inclusion of nursing diagnoses. (When nursing diagnoses were first introduced into the clinical scene, there was much consternation over the use of the term, “diagnosis.” Some physicians believed that they and they alone could use that term.) The committee also perceived a weak connection between nursing diagnoses and patient needs or resource use. They stated that if nursing diagnoses could be used to define severity of illness and the duplicate data from the UHDDS eliminated it would be helpful. During the waiting period Werley communicated with the Chief of Staff for nursing at the Veterans Administration San Diego healthcare system, Ron Norby. Although he expressed some interest in the data set he found it to be highly cumbersome. He believed that it would require labor-intensive data collection and decided not to mandate its use. (1) It was also rejected by the government.
Werley did succeed in getting some financial support from the Hospital Corporation of America foundation and two local Milwaukee acute-care facilities for research to support the NMDS and several projects followed. (1) There were disputes over ownership, however the realization that the elements in the data set were common political representations of nursing led to it being open access. Unfortunately, existing practice conditions received scant attention in the development of the NMDS. This resulted in the elements in the final NMDS representing how the individuals who represented academic and administrative backgrounds believed nursing practice should be structured, not necessarily how it existed. This hindered its acceptance into practice. (1)
The establishment of DRGs in 1983, (8) although before the formal NMDS was established, created an atmosphere in which the NMDS might have been accepted because the healthcare industry was looking for ways to meet the information needs of government payers. (1) The latter were also looking for ways to standardize information collected for billing purposes. However, the atmosphere created by the financial upheaval with the introduction of DRGs along with the fact that the NMDS was more a post-service evaluation tool than a real-time clinical tool closed the door to its acceptance by the government.
Although the NMDS was both a product of its time, ahead of its time, and subject to the policies of its developers about what nursing practice should be, the fact remains that nursing is today still very unrepresented in any data collected for many purposes including creating healthcare policy. Today the push seems to be for a standardized terminology that is interdisciplinary. This is an admirable goal. However, if one looks at the other healthcare professions one sees that they all are revenue generating while nursing remains invisible and billed in most instances as part of room and board. This deprives nursing of control of its own practice, something that most other healthcare professions enjoy.
For a more detailed look at the development and politics of the NMDS, it is suggested that one download and read Hobbs J. Political dreams, Practical Boundaries: The Case of the Nursing Minimum Data Set, 1983-1990. Nurs Hist Rev. 2011;19:127-55. Using memos, papers, letters, meeting minutes, and conference proceedings from the perspective of Harriet Werley and Normal Lang this book chapter examines the pre-conference planning, actual conference, and post-conference debates surrounding the creation of the NMDS. Much of the information in this page was obtained from that article.
References
1 Hobbs J. Political dreams, practical boundaries: the case of the Nursing Minimum Data Set, 1983-1990. Nurs Hist Rev; 2011. p. 127-55.
2 Ryan P, Delaney C. Nursing Minimum Data Set. In: Fitzpatrick JJ, editor. Annu Rev Nurs Res: Springer Publishing Company; 1995. p. 169-1494.
3 Werley HH, Devine EC, Zom CR, Ryan P, Westra BL. The Nursing Minimum Data Set: Abstraction Tool for Standardized, Comparable, Essential Data. Am J Public Health; 1991. p. 421-6.
4 Werley HH, Devine EC, Zorn CR. The Nursing Minimum Data Set (NMDS) Data Collection Manual. Milwaukee Wisconsin: University of Wisconsin-Milwaukee School of Nursing; 1988.
5 Welton J, Halloran EJ, Zone-Smith L. Nursing intensity: In the footsteps of John Thompson. In: Park HA, Murray P, Delaney C, editors. Consumer Centered Computer-Supported Care For Healthy People. Amsterdam, Netherlands: IOS Press; 2006. p. 367-71.
6 Halloran EJ, Kiley M, M. E. Nursing diagnosis, DRGs, and length of stay. Nursing diagnosis, DRGs, and length of stay. 1988;1(1):22-6.
7 Knauf RA, Ballard K, Mossman PN, Lichtig LK. Nursing Cost by DRG: Nursing Intensity Weights. Policy, Politics, & Nursing Practice. 2006;7(4):281-9.
8 AHIMA. The Evolution of DRGs (Updated). 2010
Nursing Management Minimum Data Set
The Nursing Management Minimum Nursing Data Set (NMMDS) is a research based data set of elements arrived at by consensus that are needed by nurse managers to plan, conduct, and evaluate nursing services.(1). The definitions for each of the 18 elements are standardized and divided into three broad categories of Environment (10 elements), Nurse Resources (4), and Financial Resources (4).(2) The original NMMDS had only 17 elements, the 18th was added recently.
The data elements are thought to be the minimum data needed at the management level to capture the value that nursing adds to an organization. The elements are designed so that data not specifically designated can be discovered by computing two different data elements. Nursing intensity can be discovered by adding staff mix and volume data. The NMMDS data elements are incorporated into the Logical Observation Identifier Names and Codes (LOINC). This makes them accessible for healthcare organizations for collecting and analyzing relevant data for nursing management.(1)
References
1 Kunkel DE, Westra BL, Hart CM, Subramanian A, Kenny S, Delaney CW. Updating and normalization of the Nursing Management Minimum Data Set element 6: patient/client accessibility. Comput Inform Nurs. 2012 Mar;30(3):134-41.
2 Huber D, Schumacher L, Delaney C. Nursing Management Minimum Data Set (NMMDS). J Nurs Adm. 1997 Apr;27(4):42-8.
International Nursing Minimum Data Set
Having similar concerns as nurses in the U.S. and recognizing the value of nursing data, a number of similar minimum data sets have been developed in Australia, Canada, Belgium, Ireland, the Netherlands, Iceland, Switzerland, and Thailand. After several years of discussion, in 2001, the International Medical Informatics Association Nursing Informatics Special Interest Group (IMIA NI SIG) Concept Representation Working Group and the International Council for Nurses (ICN) cosponsored a project to develop the International Nursing Minimum Data Set (i-NMDS) (1). As envisioned by the founders of the I-NMDS it would contain elements in the three areas of the USA-NMDS: setting, patient demographics, and nursing care2. Unfortunately, progress in the development is on hold, due to the size of the job and the volunteer nature of the ICN. Currently, each country has its own NMDS, designed to meet the information needs of that country. In some countries, the NMDS is related to their diagnosis related groups (DRG) which in the US is used for billing purposes. Description of the i-NMDS.
References
1 International Council of Nurses. International Nursing Minimum Data Set (I-NMDS). 2011 January
2 ICNP Collaborations. ICNP® Catalogues. 2010 July 12
Canadian Nursing Minimum Data Set
The Canadians, like their neighbors the Americans, have the situation in which health data that is collected is representative of only medical practice. In the 1990s and with a strong belief that nursing data is also important in the centralized health data system then being planned by the Canadian Institute of Health Information (CIHI), they started work on the Canadian Minimum Nursing Data Set. The purpose was to make available accessibility to standardized nursing data.1 At a 1997 working group meeting, Canadian nurses “…reached national consensus on five nursing care data elements: client status, nursing intervention, client outcomes, primary nurse identifier, and nursing intensity” to comprise their nursing minimum data Set.1
Additional Resources for Nursing Minimum Data Sets
Hobbs J. Political dreams, practical boundaries: the case of the Nursing Minimum Data Set, 1983-1990. Nurs Hist Rev; 2011. p. 127-55 doi: 10.1891/1062-8061.19.127
The above resource provides a complete history of the development of the US Nursing Minimum Data Set including its original aims, the controversies, and disappointments.
Hobbs, J. (2011).
University of Miinnesota Center for Nursing Informatics. USA NMDS.
Description of the US Nursing Minimum Data Set including why it is important.
Werley HH, Devine EC, Zom CR, Ryan P, Westra BL. The Nursing Minimum Data Set: Abstraction Tool for Standardized, Comparable, Essential Data. Am J Public Health. 1991;81(4):421-6.
The above article, with the prime architect of the NMDS as one of the authors, describes the purposes of, each element of, benefits of, implications, and the type of research that the NMDS would make possible.
Saba, V. & Feeg, V. Clinical Care Classification
University of Minnesota Center for Nursing Informatics. USA NMMDS.
The above article provides a description of the NMMDS (USA Nursing Minimum Data Set.Description of the Nursing Managment Minimum Data Set including its importance and elements of the 2006 version.
A nursing management minimum data set (NMMDS). 1993.
A description of the application for funding to develop the NMMDS.
Goossen WTF, Epping PJMM, Feuth T, Dassem TWN, Hassman A, VacDenHeuvel WJA. A comparison of nursing minimal data sets. Journal of the American Medical Informatics Association. 1998;5(2):152-63.
Although old, the above article provides information on the different uses for nursing minimum data sets in various countries and for different purposes.There is also a discussion of pluses and minuses for minimum data sets.
Created July 28, 2014
