I have finally reached the literature review process for my dissertation.
To remind the reader of this dissertation and its purpose, it is focused on the use of GIS in medical care, especially managed care. As a statistician working in a managed care quality improvement program in Colorado, I evaluated about 40,000 members enrolled in Medicaid per year, 1500 enrolled in Medicare, and 4000 to 6500 employees enrolled in the employee health program.
As a part of this activity, I was required to engage in the annual HEDIS review, which consisted of about 15 to 20 research topics (medical conditions or procedures/practices) that had to be manually looked up in the medical records, and another 20 to 25 quality improvement activities (generally an NCQA trademarked term) or performance improvement projects (the term for MCD/ MCR studies) devoted to special research topics (Chlamydia, Diabetes, Asthma care, etc., and one or two per year of the agency’s choice).
A number of these projects involved an assessment of the quality of care process, to determine if barriers existed for meeting certain healthcare goals. In theory, if those barriers were eliminated, the performance outcomes will demonstrate statistically significant improvement. These statistically significant improvements were demonstrated using standard Chi Squared or t-Test processes, and occasional Pearson’s Coefficient evaluations. GIS processes were never utilized.
My first conclusion about this program by the end of the third month was that a population health approach had to be taken. I therefore developed a population pyramid producing data pull process for the Dell/Perot System UNIX station available to analysts. This included evaluations of standard demographics information for a total of 250,000 to 500,000 individuals, depending upon the year, and those with service codes for MCD, MCR, CHP and the numerous Employee Health programs provided by this agency. I then applied these population age-gender study results to the study populations evaluated for each quality improvement study, and some of the former HEDIS defined (inactive, but still researched) QIA projects implemented years earlier.
In developing the Childhood Immunization program PIP, I added Well Visit data to the evaluation process, to determine which well visits were most linked to completion of immunizations by the age of 2 and 5 (two HEDIS measures, and two required NCQA evaluations, performed for each of the three or four programs (if you include CHP)).
I also developed a way to evaluate which clinics had the best and the worst performance. From this point on, I was engaged in a spatial analytic process of sort, evaluating urban area clinic performance, for clinics that had demographic features differing from each other. But the use of GIS to engage in this process was only imaginative.
GIS, then, was still at the ArcView GIS level, and PC data limits were at about 10-20 MB for a very expensive high-performing system; a typical system usually was able to manage a 8-10 MB storage, enough for GIS but still quite limiting.
This dissertation serves to develop a process that can be implemented for a managed care system, based on this experience in the office setting of a typical non-SAS friendly managed care quality improvement office setting.
There are several spatial analysis processes that can be applied to managed care–
- a crude mapping process using Excel and Excel extensions
- a crude process for mapping using a program that is one step up from Excel, such as Tableau
- the standard SAS-GIS (which is not user friendly, with or without SAS Enterprise capability),
- a process I developed which was 3D modeling of data in SAS, using just SAS-GRAPH, and not requiring GIS
- ArcGIS (ESRI) generated spatial analyses
- The CDC/NIH spatial analytic tool(s), which are difficult to produce maps with, but strong in terms of their analytic strengths and speed of processing
- Any of several other GIS tools now being promoted on the internet for this use (but not texted by myself)
A number of articles will be reviewed, and recommended for those interested in exploring this possible direction as part of the managed care system. As they are reviewed, they will be added to the following list:
Cresswell, K., & Sheikh, A. (2013). Organizational issues in the implementation and adoption of health information technology innovations: An interpretative review. International Journal of Medical Informatics, 82, e73-e86. doi:10.1016/j.ijmedinf.2012.10.007 Link: CresswellandSheikh_2012_HITBarriers_Int J Med Inform
Kruse, C. S., Regier, V., Rheinboldt, K. T. (2014). Barriers over time to full implementation of health information exchange in the United States. JMIR Medical Informatics, 2(2), e26. doi:10.2196/medinform.3625 Available at http://medinform.jmir.org/2014/2/e26/ Link: Barriers_medinform_v2i2e26
Mair, F. A., May, C., O’Donnell, A., Finch, T., Sullivan, F., & Murray, F. (2012). Factors that promote or inhibit the implementation of e-health systems: an explanatory systematic review. Bulletin of the World Health Organization, 90, 357-364. doi:10.2471/BLT.11.099424 Link: FactorsthatpromoteofinhibitHIT