The study identified six recommendations for HRSA and other organizations to consider as they attempt to identify facilities with critical shortages of RNs accurately and reliably. Several of these recommendations are presented below.

1. Of the methods examined in this study, the Preferred Method outlined in this report is the best choice for assessing the severity of nursing shortages in counties in the U.S. It meets more of the desirable criteria identified by the study advisory panels and it can be implemented with currently available data. Additional steps outlined below could further improve the effectiveness of this method.
2. Additional review and validation of the Preferred Method is required by stakeholders who would be affected by its implementation. Ideally, this validation should take place in a representative sample of states, counties, and facilities across the U.S., and would address the following kinds of questions:
   • Are facilities and counties classified correctly by the method? Is the method biased in favor of or against a type of facility, type of community or county, or region of the country? If so, how should the bias be addressed or overcome?
   • Are the basic data required to support the method both available and accurate for all regions and states in the U.S.? How should sampling errors for small rural counties be addressed?
   • How should facilities that have nursing shortages primarily due to persistent poor management be dealt with in the method? What criteria should be used to identify facilities with poor management and should their identities be made public?
   • Should the method be supplemented by some sort of appeals process to permit a facility with a genuine shortage to qualify for NELRP and NSSP even though the method does not place it in a sufficiently severe shortage category?
   • Should the method identify just enough “severe shortage” counties and facilities to allocate all NELRP and NSSP recipients and other related funds based on nursing shortages? Or should it identify extra facilities to provide flexibility to account for other factors?
3. More accurate estimates of RN employment and supply should be developed at the county level. This may not require new data collection if appropriate refinements can be made to the sampling frames for existing datasets, especially the NSSRN.
4. More research should be conducted on factors related to the demand for RNs, including HMO penetration, alternate service delivery models, the use of LPNs and other types of staff, and new diagnostic and treatment technologies. Factor analysis may be a fruitful avenue for additional research. Another promising avenue for research will open up when the revised HRSA Nursing Demand Model becomes available sometime in 2007.
5. More research should be conducted on factors related to the supply of RNs including RN commuting patterns, how very rural communities can recruit and retain RNs, how inner-city facilities can recruit and retain RNs, etc. The revised HRSA Nursing Supply Model is expected to be available sometime in 2007 and offer expanded research opportunities.
6. Because shortcomings in available data and extenuating circumstances might cause certain facilities to be assigned the wrong shortage designation, a formal protocol by which facilities can appeal and correct their shortage designation should be developed. The development process should consider a variety of appeal options including single facility designation changes and blanket designation changes for entire classes of facilities.