Health Human Resources Modeling: Challenging the Past, Creating the Future

by Linda O'Brien-Pallas | Oct 31, 2007

Key Implications for Decision Makers


  • Health human resources planning must be needs-based and outcomes-directed.
  • Health human resources planning should not assume that healthcare needs in the population remain constant by age and sex, even for forecasting models that focus on short 10-year planning horizons. These results clearly show that the planning models that assume future need for health human resources can be estimated solely on the basis of projected size and age-sex distribution of the population are based on faulty assumptions.
  • Changes in population health needs over time are complex. The changes observed in the level and age progression of need for health services by age varied depending on indicator of need used. Some types of needs are decreasing while others are increasing. Patterns of change in health needs vary by sex, age group, education and place of residence.
  • Consistently measured indicators of health needs that are systematically collected through period population health surveys are required to model changes in population health needs over time. Ongoing changes in the content, question wording and coding of population health surveys dramatically limited the range of health need indicators that can be used to guide planning.


  • Levels of employment of other hospital staff was significantly associated with nursing productivity, suggesting that the required number of nurses will depend on other hospital inputs.
  • In the case of nursing inpatient care over a three-to-four-year period, the rate of service output has changed, with both the direction and rate of change differing among provinces.


  • Individual, job and employer characteristics lend insight into nurses’ career intentions.
  • One size fits all retention strategies may not be preferred by nurses along different career paths, in different jurisdictions and of various ages. Policy initiatives need to be tailored to re-attract former nurses and to retain current nurses.

Executive Summary

Governments and managers are challenged to ensure that adequate and efficient nursing services are delivered to meet the health needs of Canadians and to support health-system goals. Concerns about nurse supply need to be analyzed with consideration of changing population health needs, the efficient delivery of health services and the workplace concerns of providers. Traditional approaches to health human resource planning have relied on applying current provider-to-population ratios to projected future populations; however, these approaches fall short as changes in population health needs and in provider productivity are not taken into account. Guided by the conceptual framework for health human resource planning developed by O’Brien-Pallas, Tomblin Murphy and Birch (2005), this program expands existing demographic-focused approaches to health human resource planning by moving beyond considerations of supply and utilization towards an examination of the broader social, political, economic, geographic and technological influences on the health system.

Three separate but related projects were undertaken to link population health needs to health human resource planning, to illustrate the value and challenges in using health human resource data to inform policy decisions on nursing productivity and to generate evidence based retention policies to guide nursing workforce sustainability. Using health survey data, project 1 explored the level, distribution and patterns of health indicators by demographic and social strata. In project 2, productivity was studied by analyzing select acute care nursing services using Management Information Systems data for nursing hours and other inputs and Discharge Abstract Database data for inpatient episodes of care and severity. Project 3 surveyed former nurses and registered nurses across six Canadian jurisdictions.

Project 1 demonstrated that, not only have years been added to life, but also life to years. The effect of age on health has changed over time. Regression results showed significant differences in the level and age progression of health status and health risks, even over a period of 11 years. For example, 65 year olds on average today can expect to be healthier, and hence have fewer healthcare needs, than 65 year olds on average 11 years ago. To assume that health needs by age remain constant is incorrect, even for forecasting models which are often limited to a 10-year time horizon. The results also revealed considerable complexity in the patterns of this change. The effects of year of birth varied by health indicator and by risk factor. In younger cohorts, rates of mortality, mobility, pain and smoking were decreasing, but chronic conditions and blood pressure were increasing. Those with less education were more likely to experience health problems. Although older cohorts showed a widening gap between those with low and high levels of education, the difference for younger cohorts remained fairly stable. Understanding the complex relationships between health, age and birth year as well as social indicators is vital to ensure accurate and efficient planning for future health human resources requirements. Direct independent measures of health and health risks at the population level are needed to adequately inform health human resource planning approaches.

Project 2 examined the role of productivity in health human resource planning. Productivity refers to service rendered per unit of time worked by the healthcare provider. Productivity does not mean that providers work more hours; rather, it means that providers generate more service per hour worked. Improvements in the productivity of health human resource provide a source of increased output in the healthcare sector. Findings indicated that, in the case of inpatient nursing care over a three-to-four-year period, the rate of service output has changed (as measured by severity-adjusted inpatient episodes of care) and that the direction and rate of change have differed between provinces. The employment levels of other staff were also significantly associated with the average productivity of nurses. As such, the required number of nurses to deliver a planned level of service (or manage a particular patient mix) will depend on the configuration of other hospital inputs (it is context specific) and on the methods of production. Plans to change other inputs (such as number of beds) must consider implications for nurse human resources requirements to achieve desired levels of services.

In Project 3, individual, job and employer characteristics that influenced job satisfaction, intent to retire early and risk of leaving the profession among registered nurses were examined. Nurses identified preferred policies that would attract and keep them in the profession longer. Appropriate workload, benefits packages, better salary, support for continuing education and improved work environment were very important policies for all nurses. However, policies need to be tailored to particular sub-groups. For instance, nurses over age 50 and those intending to retire early highly valued managerial support. Ontario nurses at risk of leaving would welcome greater availability of the types of positions they were seeking. Nurses under age 35 in Saskatchewan and the Atlantic region valued full-time employment. Former nurses reported leaving the profession as a result of poor work environments, personal health issues and career opportunities outside of nursing. Almost one-third of former nurses continued to work in healthcare. Policies that would attract former nurses back into nursing included appropriate workload, better salary and improved work environment. Those under age 35 particularly valued full-time employment whereas those in mid-career prioritized workplace safety.

Policy makers can improve estimates of required health human resource by incorporating population health needs, productivity analyses and evidenced based policy strategies tailored for providers. Models, practices and strategies for health human resource planning that are needs-based, outcome-directed and that recognize the complex and dynamic nature of the factors that impact these decisions need to be supported. To do so requires partnerships, analytical capacity, ability to access and link data with sustainable infrastructure, as well as ongoing evaluation to determine how changes in system delivery and in roles for healthcare providers influence health, system and provider outcomes.