Journal of Nursing Administration:
Achieving "Meaningful Use" of Electronic Health Records Through the Integration of the Nursing Management Minimum Data Set
Westra, Bonnie L. PhD, RN, FAAN; Subramanian, Amarnath MD, MS; Hart, Colleen M. MS, RN; Matney, Susan A. MS, RN-C; Wilson, Patricia S. RT(R), CPC, PMP; Huff, Stanley M. MD; Huber, Diane L. PhD, RN, FAAN, NEA-BC; Delaney, Connie W. PhD, RN, FAAN, FACMI
Authors' Affiliations: Assistant Professor (Dr Westra), Professor and Dean (Dr Delaney), Doctoral Student (Ms Hart), School of Nursing, University of Minnesota, Minneapolis; Medical Director (Dr Subramanian), Department of Pathology, Health Partners, Bloomington, Minnesota; Doctoral Student (Ms Matney), Office of the Associate VP for Health Sciences Information Technology, University of Utah, Salt Lake City; Senior Content Engineer (Ms Wilson), 3M Health Information Systems Incorporated, Murray, Utah; Clinical Professor Biomedical Informatics (Dr Huff), Intermountain Health Care, Salt Lake City, Utah; Professor (Dr Huber), College of Nursing, University of Iowa, Iowa City.
Corresponding author: Dr Westra, University of Minnesota, School of Nursing, WDH 5-140, 308 Harvard St SE, Minneapolis, MN 55455 (firstname.lastname@example.org).
Objective: To update the definitions and measures for the Nursing Management Minimum Data Set (NMMDS).
Background: Meaningful use of electronic health records includes reuse of the data for quality improvement. Nursing management data are essential to explain variances in outcomes. The NMMDS is a research-based minimum set of essential standardized management data useful to support nursing management and administrative decisions for quality improvement.
Methods: The NMMDS data elements, definitions, and measures were updated and normalized to current national standards and mapped to LOINC (Logical Observation Identifier Names and Codes), a federally recognized standardized data set for public dissemination.
Results: The first 3 NMMDS data elements were updated, mapped to LOINC, and publicly disseminated.
Conclusions: Widespread use of the NMMDS could reduce administrative burden and enhance the meaningful use of healthcare data by ensuring that nursing relevant contextual data are available to improve outcomes and safety measurement for research and quality improvement in and across healthcare organizations.
The anticipated cost savings associated with healthcare reform are in part predicated on the assumption that meaningful use of electronic health records (EHRs) can streamline care processes and increase the reuse of clinical and administrative data to improve patient safety and outcomes and increase access to care. Beginning in October 2010, the Centers for Medicare and Medicaid Services will provide Medicare incentive payments to hospitals and providers who meet the criteria for meaningful use of EHRs, and reimbursement will decrease in 2015 for those who do not meet the criteria.1 The meaningful-use criteria include electronic documentation of care and exchange of data across organizations as well as reuse of the data for quality improvement. The single most important resource in reforming the healthcare system is the need for accurate, representative, and relevant data regarding information pertaining to patient needs, care provided, outcomes realized, and information about the appropriate use of resources influencing care. Given that nurses constitute the largest group of healthcare professionals in the United States,2 it is vital that appropriate nursing clinical and contextual information is captured, stored, and linked with other healthcare data to evaluate and continuously shape ongoing system changes.
The process of quality improvement is shifting from review of paper charts to reuse of data from data warehouses, which contain extracts of data from computerized systems such as billing and claims data and, more recently, from EHRs. However, management and administrative data that describe the context of care and care delivery are missing in these repositories. Organizational variables, provider and workforce characteristics, and financial data that represent the context of nursing care influence the effectiveness of care delivery and patient outcomes.3 Nursing management data are collected in every healthcare setting; however, if the data are not captured in data warehouses and/or lack consistency in definitions and coding, it is impossible to reuse these management data to compare patient outcomes and nursing workforce issues within and across settings. These data need to be standardized and included in data warehouses along with EHR clinical data to meet the criteria for "meaningful use" of EHRs.
The Nursing Management Minimum Data Set (NMMDS) is a research-based minimum set of essential data elements that can fill the void in data warehouses to describe the management of nursing care.4 The NMMDS was developed over 10 years ago and is available as a paper-based survey upon request from the developers. With the increased emphasis on quality improvement through reuse of EHR data, it is essential to update and publicly distribute the standardized NMMDS data elements, definitions, measures, and codes to complement EHR data. The distribution of the NMMDS is best accomplished by linking it to a federally accepted national terminology that is publicly available. The Logical Observation Identifier Names and Codes (LOINC) system is one such standard with a history of incorporating survey instruments. In this article, the investigators describe the methods and outcomes of the initial steps to update the definitions and measures for 3 of the 18 NMMDS data elements, normalize these measures to current national standards, and disseminate the data set by linking the NMMDS to LOINC.
Nursing Management Minimum Data Set
The NMMDS is a research-based minimum set of essential data elements for capturing unit- or service-level nursing management data that are accurate, reliable, and useful for management decision making. It is composed of 18 data elements organized in 3 categories: environment, nursing care, and financial resources, as shown in Table 1. Each NMMDS data element is operationalized by more specific subconcepts and measures that can be linked with nursing management data already collected.
The NMMDS development initially began in 1989. Donabedian's structure, process, and outcome framework5; the Iowa Model of Nursing Administration; and the USA Nursing Minimum Data Set served as conceptual foundations for the data set. Multiple studies, one of which was supported by the American Organization of Nurse Executives, were conducted to develop and establish validity of the NMMDS data elements and definitions. Validity was established first in acute-care settings and then in long-term-care settings, ambulatory clinics, and community settings.5 In 1998, The American Nurses Association recognized the NMMDS as 1 of 2 data sets and 10 terminologies for nursing.
The value of the NMMDS is that it identifies nursing management variables that can be combined with billing and clinical data to build a better understanding on how nursing resources and the context of care influence patient safety and other outcomes. Moreover, the NMMDS can foster an increased understanding of the nursing workforce needs in terms of quantity and level of expertise specific to specialties and settings of care. The NMMDS has not been implemented in its entirety within the United States for comparison of nursing management data across settings or extensively included in data warehouses. Consequently, access to these data to support quality improvement activities or research is minimal. There is beginning research on incorporating nursing management in data warehouses.6 Specific elements of the NMMDS have proven valuable for understanding costs of care,7 impact of staff turnover,8 adverse events,9 and patient morbidity and mortality.10,11 Many of the NMMDS variables have been incorporated into the Magnet Recognition Program®, the National Database of Nursing Quality Indicators, the National Quality Forum, and The Joint Commission quality indicators. However, these efforts are focused primarily on acute-care settings. The NMMDS, on the other hand, has a broader application as it is designed to be used in any healthcare setting. There is a need to update and harmonize the NMMDS data elements with current national nursing quality efforts, health data standards, and research as well as to disseminate the results through a publicly available tool. The LOINC was chosen as the national data standard to link and disseminate the NMMDS data elements because data structures are similar, and as mentioned previously, the LOINC has been specifically used to incorporate survey instruments.
Logical Observation Identifier Names and Codes
The LOINC terminology is a publicly available, no-cost database that provides a set of universal names and codes with a similar structure to the NMMDS. The structure of LOINC includes a name-value pair, equivalent to a question or observation requiring the user to record an answer. LOINC can be used in computer databases and provides a national structure for transmitting data in electronic messages.12 LOINC was developed through funding by the National Library of Medicine and the Agency for Healthcare Policy and Research beginning in 1994 at the Regenstrief Institute, a research foundation affiliated with the Indiana University School of Medicine.13 Major goals of LOINC are to create user-friendly categories of terms, definitions, and codes that are universally used by all information systems to facilitate data exchange and use within and across healthcare organizations. LOINC is recognized by the American Nurses Association14 and the US Departments of Health and Human Services as a uniform standard for the electronic exchange of clinical health information and adopted by the National Committee on Vital Statistics for electronic exchange of laboratory results.15
The first 3 data elements of the 2005 version of the NMMDS were evaluated for (1) usefulness, (2) logical organization, (3) consistency with health data standards and research, (4) clarity of conceptual and operational definitions, and (5) the data structure for linking with LOINC. An iterative process was used to evaluate each data element. Existing standards and the literature were reviewed for conceptual and operational definitions. A resulting list of resources was compiled, and recommendations presented to the research team for consensus on the final definitions. Each NMMDS data element, subconcept, and measure was entered in Excel, and a proposed LOINC coding was developed. A small group from the national LOINC committee reviewed the definitions and the proposed LOINC coding before presenting the final updated NMMDS data elements, definitions, and coding to the full national LOINC committee for approval. Once approved, the revised NMMDS data elements with LOINC codes were incorporated into the next release of LOINC for public distribution and the next version of the NMMDS.
Results are reported separately for each of the NMMDS data elements with examples of the measures. The full list of measures is available on the University of Minnesota School of Nursing's International Classification of Nursing Practice Center for Nursing Minimum Data Set Knowledge Discovery under "USA NMMDS Updates" (http://www.nursing.umn.edu/ICNP/USANMDS/home.html).
NMMDS 01: Unit/Service Unique Identifier
The Unit/service unique identifier was defined in 2005 as the unique name, identifier, payment and geographic data for a center of excellence, service program, cluster by level of care, service/product line, or service/area where the majority of patient/client care is delivered; this is the first level of data aggregation beyond the patient/client care provider and included 9 subconcepts. The original subconcepts were unique facility identifier, unique service identifier, unique service name, unique unit identifier, unique unit name, Medicare payment category, geographic location, postal location, and country code. Of the original subconcepts, 3 were retained but updated, 6 were retired, and 3 new subconcepts added for a total of 6 subconcepts in the updated version. The unique facility identifier, geographical location, and postal code were retained and updated. Existing governmental standards were used to provide measures for these, and the coding available from the government Web sites is referenced so that as the codes change, the measures for the NMMDS data also are updated, supporting consistency in data elements and coding over time. The place of service also includes 2 for "stores" and "voluntary health or charitable agencies" that are not in the national governmental standards; these are important to include as these are places where nurses practice. Software vendors and health information technology staff track or receive notices about changes in government standards so they can continuously update their software; thus, the NMMDS coding also is updated simultaneously within systems that use these variables. Two new subconcepts were added: reporting period and facility name. The reporting period can be any 2 dates during which data are collected, that is, monthly, quarterly, or annually. Previously, dates for reporting NMMDS data were not included, nor was the name of the facility. A comparison between the previous 2005 version and the 2009 version is shown in Table 2.
NMMDS 02: Nursing Delivery Unit or Service
The nursing delivery unit or service can be any service program (product line) or physical area where care is delivered. It is the first level of aggregation beyond the individual patient/client care provider.16 The subconcepts in the 2005 version contained 37 names for types of units or services; these codes were retired and replaced with codes included in the National Database for Nursing Quality Indicators (https://www.nursingquality.org/Documents/Public/APPENDIX%20D.pdf). Mapping to an existing national standard used in 1,500 hospitals allows comparison of data collected by any other nursing setting that uses the NMMDS. The 4 subconcepts for the service and unit identifiers and names originally included in the NMMDS 01 data element were combined into 2 measures for the unique unit/service identifier and name. Table 3 shows a comparison between the 2005 and 2009 version for NMMDS 02: nursing delivery unit or service data element.
NMMDS 03: Patient or Client Population
The NMMDS data element 03: patient or client population describes the characteristics of the population served by a nursing delivery unit or service. Originally, there were 4 subconcepts for this data element: specialty, developmental focus, interaction focus, and population focus. Of these 4 subconcepts, 1 was retired, 2 were retained with new names and updated measures, and 1 subconcept was moved to another NMMDS data element. When we compared the original measures for the population specialty, the measures were redundant with measures describing the type of unit in the previous data element; hence, this subconcept was retired. There were 2 subconcepts that were renamed for clarity. Developmental focus was renamed chronological age, and population focus was renamed catchment area. Development stages were used to measure the percentage of patients served on a unit by the developmental focus; however, developmental stages have changed over time, were not sufficiently detailed to describe the age of the population served, and included overlapping groups. Measures for chronological age were changed to 5-year incremental age categories plus "fetal" and ages "1-28 days." There is no national standard for grouping patients by age. The National Cancer Institute's age grouping was selected as it provided the smallest increments for age that would be applicable across any unit or service (http://www.seer.cancer.gov/stdpopulations/stdpop.19ages.html). To prevent redundancy in the NMMDS measures, the type of client served was moved to the NMMDS 04: volume of nursing care. The volume of nursing care includes calculations for hours of care by type of client, type of nurse provider, and type of encounter. Finally, a new subconcept to capture the total population served during a reporting period was added for comparison of the size of a unit or service and becomes the denominator when calculating percentages of clients served by age or catchment area. The 2005 and 2009 version of the NMMDS 03: patient or client population data element, subconcepts, and measures are shown in Table 4.
During the initial phase of this study, the investigators examined the usefulness, clarity, and consistency of conceptual and operational definitions with governmental and health data standards and research, logical organization, and the data structure requirements for linking the NMMDS with LOINC. The first 3 NMMDS data elements were reworded, reorganized, redefined, and harmonized with existing governmental and nursing quality improvement standards and research. Now that the first 3 NMMDS data elements have been updated and are publicly available, these data elements can be used to support multilevel and multiagency analyses of the context of nursing care on patient safety, outcomes, and the nursing workforce information requirements. Given this, there are several implications for nurse managers in use of the updated NMMDS data elements.
The NMMDS includes 18 essential data elements; this study presents an update for the first 3 data elements. As identified in Table 1, implementation of the NMMDS is useful to compare the impact of various nursing care delivery models or types and amount of staffing on workforce outcomes such as staff autonomy, retention, turnover, and satisfaction; the effective of changes such as implantation of EHRs on decision making and patient safety; or the impact of staff education, certification, and facility accreditation on patient outcomes and cost savings at the unit or service level. Imagine the information nurse managers and administrators would have at their fingertips if these variables were defined, coded, and routinely collected in a standardized manner for nurse managers and administrators to add to a data warehouse for comparison of clinical and workforce outcomes. For instance, studies demonstrated that certified and advanced practice nurses improve outcomes and reduce costs for specific patient populations including mental healthcare,17 cardiac,18 neurological,19 and orthopedic conditions.20 However, these studies are limited primarily to a small sample size and are costly to conduct. In 2009, the Wound, Ostomy, and Continence Nursing Society provided a grant to Westra, Bliss, and Savik (2009) for $200,000 to evaluate the effect of certified wound, ostomy, and continence nurses on a national sample of approximately 1 million patients for outcomes of urinary and bowel incontinence, urinary tract infections, and wounds including pressure ulcers, stasis ulcers, and surgical wound. This study reuses standardized EHR and administrative data. If new data were collected with a conservative estimate of $1 per patient, the study would cost $1 million instead of $200,000. The cost of this study is possible only because home care agencies collect standardized assessment data and also track nurse visits with an associated staff ID, which can be linked with staffing characteristics such as certification. Reuse of standardized EHR clinical data along with nursing management data is critical to provide nurse managers with cost-effective information they need for management decisions.
Nurse managers, administrators, and researchers must advocate for inclusion and use of the NMMDS data definitions and coding in health information systems. The NMMDS variables are a first step in standardizing nursing management data that can explain variance in patient outcomes and factors influencing the nursing workforce. Practical steps include comparing the definitions for data elements reported in this article with existing data collected by the healthcare organization. Where data are comparable, no changes are required except to request that the data be abstracted and linked to clinical and billing data in data warehouses. If the organizational data are not comparable, nurse managers need to reevaluate the way in which they define, capture, and store the data and request appropriate changes for future comparison of nursing management data across organizations.
Once nursing management data are standardized, stored, and linked in data warehouses, new reports can be requested to understand the relationship between nursing management data with interventions and patient outcomes. Two of the quality indicators for meaningful use of EHRs include the percentage of patients receiving counseling for smoking cessation and of diabetics with adequate long-term glucose control.21 Nursing management data can extend an understanding of factors influencing compliance with these quality indicators by asking such questions as: Does an increase in compliance with these quality indicators differ by the nursing unit? With the consistent definition and coding of nursing units, comparisons can be made across hospitals affiliated with the same health system or across health systems. Additional NMMDS variables (which are in the process of being updated) include certification of nurses, staff mix, and staff turnover. How do these variables influence compliance with quality indicators, and, hence, reimbursement from Medicare in the future?
The new criteria for meaningful use of EHRs will impact financial incentives beginning in October 2010 and disincentives beginning in 2015. Included in these criteria is reuse of EHR data for quality improvement. Nursing management data, in addition to EHR data, are essential to explain variances in the quality of care. In this article, we described the process of updating the first 3 NMMDS data elements. These data elements are now available publicly through the University of Minnesota School of Nursing Minimum Data Set Knowledge Discovery Web site and distributed beginning with release 2.24 of LOINC. We anticipate that this work will help organizations incorporate management data into their quality improvement programs.
2. US Department of Labor Bureau of Labor Statistics. Occupational Outlook Handbook, 2010-2011 Edition: Registered Nurses
. Updated 2009. Available at http://www.bls.gov/oco/ocos083.htm
. Accessed January 12, 2010.
3. Maas ML, Delaney C, Huber D. Nursing outcomes accountability. Contextual variables and assessment of the outcome effects of nursing interventions. Outcomes Manage Nurs Pract
4. Huber D, Delaney C. The American Organization of Nurse Executives (AONE) research column. The Nursing Management Minimum Data Set. Appl Nurs Res
5. Huber D, Schumacher L, Delaney C. Nursing Management Minimum Data Set (NMMDS). J Nurs Adm
6. Junttila K, Meretoja R, Seppal A, Tolppanen E, Ala-Nikkola T, Silvennoinen L. Data warehouse approach to nursing management. J Nurs Manag
7. Titler M, Dochterman J, Kim T, et al Cost of care for seniors hospitalized for hip fracture and related procedures. Nurs Outlook
8. O'Brien-Pallas L, Duffield C, Hayes L. Do we really understand how to retain nurses? J Nurs Manag
9. Fogarty GJ, McKeon CM. Patient safety during medication administration: the influence of organizational and individual variables on unsafe work practices and medication errors. Ergonomics
10. Aiken LH, Clarke SP, Sloane DM, Sochalski J, Silber JH. Hospital nurse staffing and patient mortality, nurse burnout, and job dissatisfaction. JAMA
11. Tourangeau AE, Doran DM, Hall LM, et al Impact of hospital nursing care on 30-day mortality for acute medical patients. J Adv Nurs
12. Bakken S, Cimino JJ, Haskell R, et al Evaluation of the clinical LOINC® (Logical Observation Identifiers, Names, and Codes) semantic structure as a terminology model for standardized assessment measures. J Am Med Inform Assoc
13. Huff SM, Rocha RA, McDonald CJ, et al Development of the Logical Observation Identifier Names and Codes (LOINC®) vocabulary. J Am Med Inform Assoc
14. Matney S, Bakken S, Huff SM. Representing nursing assessments in clinical information systems using the Logical Observation Identifiers, Names, and Codes database. J Biomed Inform
15. Aspden P, Corrigan JM, Wolcott J, Erickson SM, eds. Patient Safety: Achieving a New Standard
. Washington, DC: The National Academy Press; 2004. Committee on Data Standards for Patient Safety, Institute of Medicine, ed.
16. Delaney CW, Huber D, eds. A Nursing Management Minimum Data Set (NMMDS): A Report of an Invitational Conference
. Chicago, IL: The American Organization of Nurse Executives; 1996.
17. Baradell JG, Bordeaux BR. Outcomes and satisfaction of patients of psychiatric clinical nurse specialists. J Am Psychiatr Nurses Assoc
18. Blue L, Lang E, McMurray JJV, et al Randomized controlled trial of specialist nurse intervention in heart failure. BMJ
19. King CA. Research reviews. The CNS's impact on process and outcomes of patients with total knee replacement. AORN J
. 2001;73(1):243, 245.
20. Russell D, Vor der Bruegge M, Burns SM. Effect of an outcomes-managed approach to care of neuroscience patients by acute care nurse practitioners. Am J Crit Care
© 2010 Lippincott Williams & Wilkins, Inc.