The comparatively simplistic provider to patient ratios and demand-based estimates cannot accurately predict need; therefore, inefficiencies in the health care system will remain. By comparison, the needs-based model provides at least the possibility of addressing most inefficiencies. This model stands out because it is essentially an iterative model that incorporates 'need' and supply data as it becomes available. Although not perfect, the model is flexible enough that it can be adjusted to include additional parameters and data to minimize the impact of unexpected developments, such as a major weather event or a severe economic downturn. Increasing the predictive accuracy of the model is also straightforward and would depend on investing into data collection efforts. As the authors discuss, the main limitation is the quality of data available; however, patient to provider ratios or demand-based models are still inferior by comparison. A common approach used for validating predictive models such as the one developed by Murphy and colleagues (2012) is to test it using historical data. While historical data may not have the same quality as today's data does, it might still be informative to see how the model performs when the real-world outcome is already known. It would be particularly interesting to test the model using the data available at the beginning of the current obesity epidemic to see how well it predicts the increase...

If the model performs well at the local level, then its efficacy could be tested under different conditions using a case study experimental design. The potential savings in health care expenses should be more than enough to justify this research. Defining the population size limit of the model may be important given the positive outcomes of interventions implemented at the institutional level; interventions that had a dramatic impact on RN retention and absenteeism rates (Murphy et al., 2012). The above recommended experimental approaches are designed to move the model out of the laboratory and into real-world settings. If this transition isn't made successfully, the value of the model remains limited to the theoretical. If the model proves reasonably robust in real-world settings, then its value for predicting human resources based on need would increase substantially, as would its value for reducing health care costs.
References

Murphy, Gail Tomblin, Birch, Stephen, MacKenzie, Adrian, Alder, Rob, Lethbridge, Lynn, and Little, Lisa. Eliminating the shortage of registered nurses in Canada: An exercise in applied needs-based planning.…