HRSA - U.S Department of Health and Human Services, Health Resources and Service Administration U.S. Department of Health and Human Services
Order Publications
Grants Find Help Service Delivery Data Health Care Concerns About HRSA
Methods for Identifying Facilities and Communities with Shortages of Nurses, Technical Report

Study Background and Context

This report summarizes the findings of the various components of this study of methods for identifying facilities and communities with shortages of RNs. It documents the strengths and weaknesses of different methods for assessing the extent of shortages of RNs. The report is presented in seven sections, each summarizing a different aspect of the study:

  • Federal Initiatives to Address Nursing Shortages
  • Initial Literature Review
  • Data Sets and Compilations
  • Methods and Analyses Based on Facility Data
  • Methods and Analyses Based on Geographic Data
  • Preferred Method
  • Study Recommendations

In addition to summarizing these research components of the study, this report presents a series of conclusions designed to inform policy makers and other researchers who may be interested in implementing or adapting one or more of these methods in the future.

A. Federal Initiatives to Address Nursing Shortages

In 2004, realizing that the current shortage designation process was too narrow in scope and that RN shortages were likely to worsen over the next 20 years, HRSA issued a Request for Proposals for a two-year research project to gather information and insights in support of the development of a new methodology for identifying health care facilities and agencies with critical shortages of RNs. The New York Center for Health Workforce Studies at SUNY Albany was selected to conduct this project.

There is growing recognition and efforts are underway to increase production of RNs and use incentives to target new graduates to facilities and agencies with the most critical shortages of RNs. However, there are issues that must be taken into account when assessing need and demand for RNs and identifying health care providers with the most critical shortages of RNs. These include:

    • Should indicators developed to measure critical shortages of RNs be based on need for RNs or demand for RNs?
    • Can standard indicators that measure critical shortages of RNs be applied to all of the eligible settings included in this project?
    • Can variations in the supply of and demand for RNs by region, geography (i.e., rural or urban), setting, or facility be accounted for in indicators that measure RN shortages?
    • Are there setting-specific data sets available at the national level that include the elements needed to measure critical shortages of RNs?
    • Can a process be developed that identifies facilities with the most serious shortages of RNs so that Federal resources can be targeted on the neediest facilities?
    • How can true shortages of RNs be distinguished from shortages created by poor management practices?

A careful review of the literature helped to inform the discussion of these and other related issues. Through the identification and review of existing methods and models for measuring health professional shortages, information on these issues will be obtained and shared with each of four expert panels, who are providing guidance for this project. It is unlikely that standard data sets on staffing will be available for all of the health care settings included in this project. Rather, data may be available for only some providers, e.g., the American Hospital Association nurse staffing data set for acute care facilities, or the Centers for Medicare and Medicaid Services (CMS) Online Survey, Certification and Reporting (OSCAR) data set for long-term care providers. The information on staffing for some types of health care providers may be less than adequate, or it may not be available at the national level.

An effective study should take all of these issues into account while researching and testing the development of a national methodology to measure shortages of RNs. Current methods are inadequate. A better method would support several government incentive programs to attract new RNs. It would also provide a better basis for monitoring RN shortages locally and nationally.

B. Study Overview

This study was conducted over a two-year period, starting in the fall of 2004, during which nine different research components were carried out. Each component is summarized in the body of this report in roughly the chronological order they were conducted during the study:

1. Project Goals and Objectives

The primary goal of this study was to conduct research on the necessary components of a comprehensive, nationwide methodology to identify facilities and communities with critical shortages of RNs across the U.S. and its territories in order to target the placement of Federally-obligated RN scholars and loan repayers. This research, which involved statistical analysis supported by expert opinion, took into account population needs, practice settings, appropriate staffing levels, and nursing education, among other aspects of the supply of and demand for RNs. As a secondary benefit, the project revealed important insights about the differences in the use and distribution of RNs across the various settings and geographic areas of the country.

Ultimately, this research will support the development of a comprehensive method for identifying the health care facilities and agencies with the most critical shortages of RNs. This will permit more effective targeting of resources to encourage service-obligated RNs to work in the facilities with greatest need.

2. Expert Advisory Panels

The study was conducted under the guidance of four expert advisory panels, one for each of four types of health care organizations: hospitals, home health agencies, nursing homes, and public health agencies. The names of the panelists can be found in Appendix B.

Project staff worked to achieve the following objectives in support of the primary goal of the study:

  • Identify and define indicators and measures that reflect critical RN shortages for the four types of facilities;
  • Assess the availability of data sets that can be used to determine RN staffing needs nationally in each of the settings listed above;
  • Develop quantifiable key measures of nursing shortages based on key indicators described above as well as the available data sets that include the necessary data to calculate the key measure.
  • Determine whether these key measures of shortage can be incorporated into a comprehensive national methodology to identify facilities and agencies with critical nursing shortages based on the following criteria:
    • the measure accurately quantifies nursing shortages in a specific health care setting;
    • the measure either can be calculated using an available national data set or the data can be collected and validated at the facility level.
  • Establish an analytic framework that can be used for a comprehensive methodology to determine critical nursing shortages across a variety of health care settings.

3. Characteristics of an Ideal Shortage Designation Method

Early in the study a number of characteristics were identified as especially desirable for any method to identify facilities and communities with shortages of RNs. These characteristics, some of which may not be attainable, included:

  • A common method to be used across the nation;
  • Ease of calculation of the RN shortage index for individual facilities and communities;
  • Implementation using existing data sets, with no additional data collection required;
  • Comparison of shortages of RNs both within and between different types of facilities;
  • Comparison of RN shortages across different states and other geographic jurisdictions;
  • Consistency of shortage severity estimates with shortage assessments by local experts;
  • Identification of shortages in facilities due to poor management; and
  • Easy updates to the method to reflect more recent conditions, situations, and relationships.

One important Federal response to the national nursing shortage was the Nurse Reinvestment Act, which was enacted in August 2002. The Act reauthorized the NELRP, which provides loan repayment to RNs in return for work at facilities or in communities with a shortage of RNs, and established the Nursing Scholarship Program. Eligible placement sites for these programs were expanded to include:

  • Ambulatory surgical centers;
  • Federally designated migrant, community public housing, or homeless health centers;
  • Federally qualified health centers;
  • Home health agencies;
  • Hospice programs;
  • Hospitals;
  • Indian Health Service centers;
  • Native Hawaiian health centers;
  • Nursing homes;
  • Rural health clinics; and
  • State or local health department clinics or skilled nursing facilities.

The method used for the identification of qualified placement sites used a combination of geographic and facility designations. In 2002, the New York Center for Health Workforce Studies assisted the Bureau of Health Professions by developing an up-to-date list of nursing shortage hospitals and counties throughout the U.S. and its territories. The Center used two separate methodologies, one to identify private, non-profit hospitals with shortages of RNs and the second to identify counties with shortages of RNs.

Because this approach relied on hospital nursing data to identify facilities with nursing shortages, it failed to quantify nursing shortages experienced by any providers except hospitals. Most of the other types of facilities included on the list above were considered categorically eligible placement sites, based on the premise that they faced critical shortage of RNs.

C. Initial Literature Review

The first component of the research involved a careful review of the literature, focusing on characteristics of RNs relevant to the task of understanding current and future shortages. The discussion that follows summarizes a variety of relevant statistics.

1. Characteristics of RNs

The two demographic characteristics most relevant to shortages of RNs were gender and age. The gender mix of RNs was important because it reflected the size of the pool of potential candidates from which to recruit new RNs.

The age distribution was important because it dictated the numbers of existing RNs who will leave nursing in the future, creating a need to replace them in the workforce.

Table 1 provides estimates of the percentages of active RNs in the U.S. by gender and age group. Although 6.1% of RNs were men in 2004, which is higher than in previous years, nursing remains a female-dominated profession. This means that, at least in the near future, recruiting more men to the profession is not likely to be an important avenue for increasing the supply. The table also reveals that by 2014 it will be necessary to recruit more than 400,000 new RNs just to replace those RNs older than age 55 who are expected to retire from active nursing practice. In fact, the latest estimates developed by the Bureau of Labor Statistics [BLS, 2006] indicate that the U.S. will require 1.2 million new RNs by 2014 to meet the nursing needs of the country, 500,000 to replace those leaving practice and an additional 700,000 new RNs to meet growing demands for nursing services.

Table 1. Active RNs in the U.S. by Gender and Age Group, 2004

Age Group Male Female Percent
< 25
25 to 29
30 to 34
35 to 39
40 to 44
45 to 49
50 to 54
55 to 59
60 to 64
65 +

Source: 2004 NSSRN

2. Employment Settings

Figure 1 shows that hospitals continued to be the major employer of RNs in 2004, although the percentage of RNs working in hospitals declined from 59.1% in 2000 to 57.4% in 2004. The percentage working in public or community health organizations declined from 18.3% in 2000 to 11.0% in 2004.

A fact hidden in these simple employment statistics was that the day-to-day demands on many of these RNs, especially those employed in hospitals, increased dramatically over the past two decades. In fact, increases in patient acuity in hospitals and nursing homes resulted in a corresponding increase in the stress of nursing practice that caused a growing number of RNs to leave active patient care.

Figure 1. RN Employment by Setting, 2000 and 2004


Source: The Registered Nurse Population, March 2000. USDHHS, 2001. 2004 NSSRN, USDHHS, 2006

3. Trends in Supply

Between 1980 and 2004, the number of active RNs in the U.S. grew by nearly 90%. In 2000, there were more than 2.4 million active RNs, an increase of more than 1.1 million since 1980.

Between 1996 and 2000, the total number of RNs grew by only 1.3% each year, compared with average annual growth of 2% to 3% in earlier and later years (Figure 2). This slowdown in growth between 1996 and 2000 was attributable to two trends: a declining number of candidates passing the RN licensing examination annually and an increasing number of RNs leaving the field [1].

This slowdown was temporary, however, as the growth in the supply of RNs resumed between 2000 and 2004, more than keeping up with the growth in the population over the same period. The number of active RNs per 100,000 population nationally decreased from 798 in 1996 to 782 in 2000 (Figure 3). There was also wide variation in RNs per 100,000 population across the country. Massachusetts and South Dakota had the highest number of employed RNs per capita in 2000, 1,194 and 1,128 per 100,000 population, respectively. California and Nevada had the smallest number of employed RNs per capita, 544 and 520, respectively [1].

The number of candidates passing the RN licensure examination decreased steadily since 1995. Between 1995 and 2001, the number of RNs passing the licensing exam declined by nearly 28% [2].

Figure 2. Number of Active U.S. RNs, 1980 - 2004


Source: USDHHS, Findings from the National Sample Survey of RNs 2000, 2004

Figure 3. Active RNs per 100,000 Population, U.S., 1980 to 2000


Sources: USDHHS, National Sample Survey of RNs, 2004 and earlier; Population Estimates Program, Population Division, U.S. Census Bureau.

The number of graduates of RN education programs in the country also declined between 1995 and 2001. While RN production grew steadily in the early 1990s, the total number of U.S.- educated candidates taking the RN licensing examination dropped between 1995 and 2001, with nearly 29% fewer RNs graduating in 2001 than in 1995 [2]. Bachelor degree RN graduates (BRN) dropped by 20% while associate degree RN graduates (AND) declined by 28%. Although RN enrollments are increasing and the numbers of RN graduates in 2002 and 2003 were higher than the number of RN graduates in 2001 [3, 4], these figures are not yet back to 1995 levels.

4. Geographic Distribution

These national estimates and projections tell only part of the story. The two maps presented on the next page provide additional perspective on the supply of RNs in the U.S. in 2004. Figure 4 shows that the geographic dispersion of active RNs in 2004 was far from uniform across the country. In fact the ratio of the highest to lowest RN per capita ratios was nearly 4:1, with the highest ratios in the District of Columbia (2,236 RNs per 100,000 population) and New Hampshire (1,321), and the lowest in California (603) and Nevada (612).

The range of ratios by county was even greater, which highlights one of the challenges for anyone interested in identifying counties or facilities with shortages of RNs. It is essential to have access to detailed data on RNs in counties in order to develop accurate estimates.

Figure 5 provides an additional perspective on this geographic variation, the change over time in the RN per capita ratios. This map shows that seven states (Connecticut, Florida, Idaho, Louisiana, Massachusetts, Maryland, and Rhode Island) experienced a decline in the number of active RNs per capita between 2000 and 2004. On the other end of the supply change spectrum were Alaska, District of Columbia, and New Hampshire, all with increases in active RNs per capita of over 25%. After discarding these three outliers, the Pearson correlation coefficient between the 2004 supply of RNs and the change in supply between 2000 and 2004 was only -0.039 (NS).

Figure 4. RNs per 100,000 Population in the U.S., 2004


Figure 5. Percent Change in RNs per 100,000 Population in the U.S., 2000 to 2004


5. Projections of Future Supply

The National Center for Health Workforce Analysis at HRSA has projected a growing shortage of RNs over the next 15 years, with a 12% shortage by 2010 and a 20% shortage by 2015 (Figure 6). The projected shortage is the result of the expected increase in demand, coupled with a relatively stable supply of RNs [6].

Figure 7 updates these projections based in part on the 2004 NSSRN. Total numbers of RNs may rise until 2016 if age-specific cohorts follow patterns observed in the RNSS between 2000 and 2004. This is in large part because the sizes of birth cohorts in nursing tend to increase well into ages 50 to 55, and so a number of baby boomers (those born between 1947 and 1964) may still enter nursing as a second career over the next 10 years.

This does not mean that problems will not be felt until after 2016, however. Using these projections of numbers of RNs,  total population, and the population age 65 and older from the U.S. Census Bureau, Figure 7 shows that the number of RNs per 100,000 population will peak in 2012, while the number of RNs per 100,000 population age 65 and older will peak in 2008 and decline by 5% (to below current rates) by 2012.

Figure 6. National Supply and Demand Projections for FTE RNs, 2000 to 2015


Source: Bureau of Health Professions, RN Supply and Demand Projections

Figure 7. Indexed Projections of RNs per 100K Pop, RNs per 100K 65+ Pop, and Projected Numbers of Active RNs, 2004 to 2024


Source: CHWS, 2006

6. Nursing Shortages

A review of the literature revealed a number of studies examining future shortages of RNs relevant to this study. Some of the key findings are summarized briefly below.

  • Health care providers across a variety of settings reported increasing difficulty recruiting and retaining RNs, particularly in hospital settings [7, 8].
  • There were indications that the attrition from clinical settings may be related to dissatisfaction with working conditions. The 2004 NSSRN asked RNs about job satisfaction and found that 76% of RNs employed by hospitals and 75% of RNs employed by nursing homes were satisfied with their jobs, compared to 82% of RNs employed in nursing education and 83% of RNs employed in occupational health. Staff RNs across all settings were less likely to be satisfied with their jobs, as were older RNs, with the exception of those employed in ambulatory care [1].
  • Experienced RNs who left clinical settings identified a variety of reasons for their decision to leave, including lack of autonomy, heavy workload, too much paperwork, lack of opportunity for professional growth, inadequate staffing, and concerns about the quality of care. In some instances, these RNs went on to become advanced practice nurses (APNs) and return to clinical settings with more skills, more autonomy, and higher wages.
  • There is increasing concern about the impact of RN shortages on the quality of health care. A growing body of evidence demonstrates that hospitals with lower ratios of RNs to patients had more adverse events than hospitals with higher RN to patient ratios [9, 10, 11].
  • Several states have passed legislation prohibiting or limiting mandatory overtime for RNs and one state passed legislation mandating minimum nurse staff ratios in hospitals and nursing homes [12].

The current shortage of RNs and concerns about future shortages have led to new efforts—including this study—to address the problem of identifying facilities and communities with shortages of RNs.

D. Data Sets

Based on suggestions from the study advisory panels, four steps were implemented to develop criteria and methods to use for identifying facilities and communities with shortages of RNs. The four steps were:

  • Designate data requirements, data elements, and data sets;
  • Acquire data sets to use in pilot analyses;
  • Perform pilot analyses for assessing different methods;
  • Document the analyses for interested stakeholders.

1. Indicators and Corresponding Data Elements

These indicators were selected for inclusion based on the extent to which they were associated with facilities and agencies that have a shortage of RNs due to factors beyond their control (e.g., being located in a geographic area with few RNs). The advisory panels identified potential indicators at both the community and facility levels.

Community Indicators provide a critical context for any nursing shortage designation process. A number of community indicators identified by the expert panels seemed particularly relevant:

  • Demographic Context
    • Rural or urban;
    • Age distribution of population;
    • Percent of population using Medicare or Medicaid;
    • Median population income; and
    • Percent of population in poverty.
  • Nursing Context
    • RNs per 100 hospital beds;
    • Local nursing wages;
    • Numbers of nursing schools and graduates; and
    • Numbers of new RNs passing the National Council Licensure Examination for Registered Nurses (NCLEX).

Facility Indicators further refine and inform the shortage designation process. Facility indicators suggested by the panels included:

  • Facility Indicators
    • Type of facility; and
    • Facility size;
  • Workforce Statistics
    • Turnover rates;
    • Vacancy rates;
    • Hard-to-fill positions;
    • Staffing ratios (e.g., RNs per 100 beds, support staff per RN);
    • Poor facility outcomes (e.g., bad outcomes per 1,000 admissions);
    • Case mix and acuity;
    • Worker satisfaction; and
    • Turnover of leadership.

2. Identification and Compilation of Data

Data were compiled for two different tracks for this study: an “ideal” shortage designation methodology that incorporates all essential indicators required to identify shortages of RNs in either facilities or communities; and a “fall back” methodology that represents the best possible solution based on currently available data.

  • Indicators for an “ideal” methodology. This step required the identification of facilities and communities that had data for all of the kinds of indicators listed above. Potential pilot sites considered by study staff were the Veterans Administration, Hospital Corporation of America, Health and Hospitals Corporation of New York City, and states such as North Dakota, North Carolina, Iowa, Pennsylvania, California, and Delaware.
  • Indicators for a “fall-back” methodology. This step involved identifying data elements from the lists above that were available for facilities and communities of all different types across the U.S.

Two important data sources were used in this project: the Survey of Nurse Employers in North Carolina conducted by the North Carolina Center for Nursing; and the Area Resource File [ARF, 2004 release]. The facility variables were obtained from the Survey of Nurse Employers in North Carolina, and the community variables were obtained from the ARF database. The number of observations used to estimate the models was 325. There were four types of facilities estimated in this study: hospitals (65), home health facilities (79), long-term care facilities (128), and public health facilities (53).

Data were also obtained for 141 facilities in North Dakota (35 hospitals, 28 home health agencies, 45 long-term care facilities, and 33 public health agencies). These data were collected by the Center for Rural Health at the University of North Dakota, using questionnaires and definitions patterned after those used in North Carolina.