Geospatial distribution of primary care physicians in relation to the health status for the selected counties in Maryland.
This study is aimed at identifying the greatest need for primary care in Prince George's County, Maryland. This was addressed by defining healthcare and other parameters of primary care need, documenting the geographic distribution of these parameters and then, based on a synthesis of these findings, identifying areas that reflect differential levels of primary care need. To place the County data in the context of the region and state, the data are presented for four additional jurisdictions: Montgomery, Anne Arundel, Baltimore, and Howard counties. These counties either border Prince George's County and/ or have similarities in population characteristics. Primary care has been defined by the Institute of Medicine as the "provision of integrated, accessible health care services by clinicians who are accountable for addressing a large majority of personal health care needs, developing sustained partnership with patients, and practicing in the context of family and community." (IOM, 1994). This definition remains viable today with its hallmark focus on the patient, family and community and with care facilitated and augmented by teams of providers working within integrated delivery systems. Integrated care includes the provision and coordination of services that address health care needs at stages throughout a patient's life cycle and continuous over time. This care focuses on disease prevention, chronic disease management and episodic care for acute systems.
The centerpiece of this assessment is on geographic mapping based on applying the Geographic Information System. A Geographic Information System, also called GIS, is a computer-based system to aid in the collection, maintenance, storage, analysis, output, and display of spatial data (Hanchette, 2003). Geospatial mapping of health data can be instrumental in visualizing patterns and generating questions that may not have otherwise occurred to researchers and the public. More recently, GIS has emerged as a new technology in public health. In particular, it provides analytical tools for health geography and epidemiological research in cases where geographical display is important. As a spatial analytical tool, GIS serves to advance the knowledge base of health geography and informatics.
The health status of a population reflects its demographic and socio-economic composition, as well as the need for and effectiveness of its health care delivery system. We obtained a limited number of hospital discharge data from the Maryland Department of Health and Mental Hygiene (DHMH). Of the available data, we selected two of the most common conditions, myocardial infarction and asthma, to provide a selected picture of the health status of Prince George's County residents, and to compare their health with the residents in surrounding jurisdictions.
There are a number of characteristics that can be used to assess need for primary care. Two were used for this assessment: population size and race/ethnicity. Race/ethnicity of populations has been associated with differential risks for disease as well. Prince George's County's population reflects a large and diverse majority African American population and is a County with the wealthiest African American population in the nation.
PRIMARY CARE PROVIDERS
Primary care providers serve as a principal point of contact for patients seeking to maintain optimum health within a health care system. Primary care physicians include medical specialists in family practice, internal medicine, pediatrics and obstetrics and gynecology.
We conducted the analysis on four selected socio-demographic factors among five jurisdictions. All data were collected at the ZIP code level and were obtained from the Census 2010 except the household income which was obtained from Census 2000. We adopted the Arcgis Desktop 10 (ESRI, 2012) for the geographical mapping. ESRI's ArcGIS Desktop 10 with ArcMap platforms has the industry recognized out-of-box spatial analysis tools and Application Programming Interfaces (APIs).
An age-adjusted rate is a weighted average of the age-specific (crude) rates, where the weights are the proportions of persons in the corresponding age groups of a standard population. The potential confounding effect of age is reduced when comparing age-adjusted rates computed using the same standard population. These include the 2010 US standard population as well as standard millions for the US population.
Two Census variables (population size, percentage of African Americans/Blacks) that can serve as surrogates for need for primary care were selected and compared across the five jurisdictions. In addition to the geographical map, a quintile ranking was used to order and compare the ZIP codes by each of four different variables. A quintile refers to one-fifth of the sample or population. A chart alongside each map displays a) the number of ZIP codes in the highest or lowest quintile, by jurisdiction, b) the percent of ZIP codes within the highest or lowest quintile, by jurisdiction, c) the number of residents associated with those ZIP codes in the highest or lowest quintile, by jurisdiction, and d) the percent of residents associated with those ZIP codes in the highest or lowest quintile, by jurisdiction. For the primary care physicians, the quintile analysis could serve as a method to identify the areas that are in need of primary care..
First, we compared Prince George's County to other jurisdictions with populations of similar size. The largest county in Maryland is Montgomery, with a population of 971,777. The next four most-populous counties include Prince George's, with a population of 863,420; Baltimore, with a population of 805,029; Anne Arundel, with a population of 537,656; and Howard, with a population of 287,085. (Note: Baltimore City and Baltimore County are separate entities and Baltimore City is not included in the county's population.) Within these five jurisdictions, Prince George's County ranked third in percentage of population residing within top-quintile ZIP codes (ZIP codes with a population greater than 313, 938). Close to half of Prince George's County residents (47.2%) are located in these 10 top-quintile ZIP codes (see Figure 1). The Maryland Census data can be obtained from the Web site (Maryland State Data Center, 2012).
The 2010 Census indicated that the largest ethnic group in Prince George's County is non-Hispanic Blacks (63.52%). This percentage is greater than the proportion of Blacks across the entire state (nearly 30% in the 2010 Census), and the proportion of African-American/Black across the entire nation (12.2%) (Kaiser Family Foundation, 2011). The geographic analysis indicated that among the five most-populous jurisdictions, Prince George's County has 29 ZIP codes in the top quintile, i.e., ZIP codes having populations with greater than 39.9% of African-Americans/Blacks. These 29 ZIP codes consisted of more than half a million residents. Baltimore County ranked second, with 6 ZIP codes in the top quintile, consisting of 84,940 residents. No other jurisdictions had more than three ZIP codes in the top quintile (see Figure 2).
TWO CASE STUDIES OF "HEALTH STATUS" USING HOSPITALIZATION DISCHARGE DATA BY JURISDICTIONS
We analyzed the health status by first computing the age-adjusted rates of myocardial infarction and asthma at the ZIP code level, using the U.S. standard population from the 2000 Census. The GIS (ESRI, 2012) was used to map these rates in order to examine the geospatial clustering of the disease incidence. These clusters are often termed "disease hot spots." The identification of these hot spots will allow investigators to focus on these areas, identify the risk factors associated with the hot spots, and implement effective health care services.
We compared age-adjusted myocardial infarction hospital discharge rates for Prince George's County with those in the neighboring Maryland jurisdictions. These rates are presented in quintiles in the geographical map by ZIP codes. In addition, we also computed the number and percent of ZIP codes in the top quintile, and the number and percent of residents in these ZIP codes for each county.
Figure 3 presents the myocardial infarction age-adjusted rates by jurisdiction and the Maryland average rate. Surprisingly, all five jurisdiction rates were higher than the Maryland state average (81.8 per 100,000 residents). The rates for these five jurisdictions ranged from 91.7 to 190.6. The rate for Prince George's County (110.0) was higher than the rate for Montgomery, but lower than the other three jurisdictions (Anne Arundel, Baltimore, and Howard). A geographical map of the age-adjusted rates per 100,000 residents was developed at the ZIP code level across five jurisdictions. The map (Figure 4) revealed that three ZIP codes in Prince George's County fell into the 5th quintile (i.e., a rate > 183.1) with 29,531 residents. This number was substantially lower than that of Anne Arundel County, where 10 ZIP codes fell into the 5th quintile with 244,317 residents (45.4% of the county population), and lower than that of Baltimore County, where 12 ZIP codes fell into the 5th quintile with 257,551 (32% of the county population).
Figure 5 presents the age-adjusted rates of asthma by jurisdiction and the Maryland average rate. Surprisingly, four of the five jurisdiction rates were higher than the Maryland state average (83.8 per 100,000 residents). The rates for these five jurisdictions ranged from 74.0 to 233.1. The rate for Prince George's County (127.4) was higher than the rate of Montgomery County (90.1) and Howard County (74), but lower than Anne Arundel County (137.2) and Baltimore County (233.1).
A geographical map of the age-adjusted rates per 100,000 residents was developed at the ZIP code level across five jurisdictions for a more detailed examination of asthma distribution. The map (Figure 6) revealed that five ZIP codes in Prince George's County fell into the 5th quintile (i.e., a rate > 170.3), with 145,851 residents. This number was substantially lower than that of Anne Arundel where seven ZIP codes fell into the 5th quintile with 207,018 residents (38.5% of the county population), and lower than that of Baltimore County where 16 ZIP codes fell into the 5th quintile with 425,190 (52.8% of the county population).
PRIMARY CARE PHYSICIAN SUPPLY ANALYSIS
In 2010, there were approximately 4,870 active non-federally employed primary care physicians practicing in Maryland. These included specialists in family practice, internal medicine, pediatrics and obstetrics and gynecology. Of this number, 2,867 practiced in the selected five jurisdictions and 465 in Prince George's County.
The rate of primary care physicians in Prince George's County was the lowest (53.9 per 100,000 residents) among five jurisdictions. This rate was 30 points lower than the average rate for the state of Maryland (84.4 per 100,000 residents). Anne Arundel County had the second-lowest rate among the five counties, 66.6 per 100,000 residents. The highest rate belonged to Baltimore County (101.2 per 100,000 residents), which was more than twice the rate of Prince George's County. Montgomery County had the second highest rate, with 94.9 per 100,000 residents (see Figure 7).
A geographical map of primary care physicians per 100,000 residents was examined for each ZIP code across the five counties for a more detailed examination of primary care physician distribution. The map (Figure X.14) revealed that 11 ZIP codes areas in Prince George's County fell into the 1st and 2nd quintile (i.e., < 9.6 primary care physicians per 100,000 residents) with 138,676 residents. This is contrasted with neighboring Montgomery County, where 16 ZIP codes were in the 1st and 2nd quintiles, with only 87,775 residents.
To gain a better understanding of how 2,867 primary care physicians were distributed across the five jurisdictions, a "density" map was created using various advanced spatial analysis methods. The spatial density of the primary care physicians was determined by the number of physicians normalized by the size of the area. The area was defined as a function of the mean nearest area distance. Geographic areas with a density higher than a selected threshold were then circled and displayed on the map. The selected areas represent clusters of primary care physicians. Figure 9 shows that while there were several clusters in each of the jurisdictions, the clusters were smaller for Prince George's County. Identifying these clusters could help investigators examine the underlying factors associated with the clustering.
Figures 10 and 11 present the geographical locations of primary care physicians, overlaid with the age-adjusted myocardial infarction rates and the age-adjusted asthma rates, respectively. These two maps visually demonstrate the demand for patient care and the supply of the primary care physician workforce. It was apparent that the clustering of primary care physicians did not correspond to the higher rates of myocardial infarction and asthma. For example, the age-adjusted rates of myocardial infarction and asthma were substantially higher than the south region of Montgomery County, but the density of primary care physicians was one of the highest among five jurisdictions.
According to a report released by the U.S. Census Bureau, 20 percent of Prince George's County's eligible population did not have any health insurance as of 2005, the highest rate in Maryland. This translates to more than 150,000 people in Prince George's County do not have health insurance, which also the highest number in the state. While roughly 76,000 of the uninsured Prince George's patients listed in the census report make less than $25,000 a year, the other half make more but chose not to have health insurance due to the costs involved (Valentine, 2008). The results indicated that the primary care physician workforce for Prince George's County was substantially less than the neighboring jurisdictions. Specifically, the rate of primary care physicians for Prince George's County was below the average rate of primary care physicians per 100,000 residents in Maryland, and below any of the four neighboring jurisdictions. When considering the lowest two quintiles as the high primary care need area, the rate of primary care physicians supply was smaller than or equal to 9.6 per 100,000; 138,677 residents (16.2% of the population) in Prince George's County lived in this area. The "overlay" geographical analysis case studies indicated that there was a disparity between the primary care physician locations and the rates of myocardial infarction and asthma hospital discharges.
These findings reinforced the previous reports conducted by Rand (Lurie, Harris, Shih, Ruder, Price, Martin, Acosta, & Blanchard, 2010), by 2010 Primary Care Needs Assessment (DHMH, 2011), and Maryland Physician Workforce Study (MHCC Extramural Report, 2011). However, the current study expanded on previous studies by including the full range of primary care healthcare workforce categories, provided an analysis at the ZIP code level, and applied geospatial mapping to investigate both the areas of high primary care need and the variations within the Prince George's County. Although the study did not attempt to investigate the reasons for these disparities, the literature suggested that economic factors in the region may influence the recruitment and retention of this professional workforce. Health care workers--particularly those providing direct services--may face many issues related to safety and work benefits in their work environments. The lack of having teaching hospitals with an academic affiliation also may contribute to this challenge of recruitment and retention. These factors should be carefully studied in future assessments of the health care workforce. They should also be considered in any attempts to understand why Prince George's County had substantially lower rates of health care workers than other comparable counties in Maryland.
The ZIP code level data analysis may provide more detailed sub-county level information, but it is not without limitations. Some ZIP codes located in the sparsely populated areas of the county have a small number of residents and are thus more likely to have no patient care professionals providing care in the area. As a result, the rate computed for such areas is zero. On the other hand, if it happens that one or a few patient care professionals do provide services in these areas, the resulting rate could be unrealistically high. These zeros and inflated rates may become "noise" for the true pattern of the rates of patient care workforce computation and for the geographical mapping.
The ZIP code level patient care workforce data analysis may be compared to the Health Professional Shortage Areas (HPSAs), Medically Underserved Areas (MUA), and Medically Underserved Populations (MUP). Currently, the HPSA data is based on the Census tracts, which do not match with the ZIP code areas, making it difficult to combine with the patient care workforce data with HPSAs without the alignment with the ZIP code. Future studies may also consider the socioeconomic index in relation to geographical mapping and the patient care workforce. Multiple density clustering-i.e., clusters of residents, clusters of patients, and clusters of the patient care workforce-may also be considered.
Cherry, K. (2007). Difference Between Psychologists and Psychiatrists. Retrieved from http://psychology.about. com/od/psychotherapy/f/psychvspsych.htm
Department of Health and Mental Hygiene. (2011). 2010 Primary Care Needs Assessment. Retrieved from http://ideha.dhmh.maryland.gov/IDEHASharedDocuments/PCO_Needs_Assessment_11_16_11.pdf
ESRI. (2012). ArcGIS Desktop Version 10. Retrieved from http://www.esri.com/software/arcgis/about/ gis-for-me.html
Hanchette, C. L. (2003). Geographic Information Systems. In P.W. O'Carroll, Y.A. Yasnoff, M.E. Ward, L.H. Ripp, and E.L. Martin (Ed.), Public Health Informatics (pp. 431-466). New York, NY: Springer.
Health Resources and Services Administration. (2011). HRSA Shortage Designation: Health Professional Shortage Areas & Medically Underserved Areas/Populations. Retrieved from http://bhpr.hrsa.gov/ shortage
Institute of Medicine (1994). Defining Primary Care: An Interim Report. Retrieved from http://www.nap. edu/openbook.php?record_id=9153&page=R1
Kaiser Family Foundation. (2011). Distribution of U.S. Population by Race and Ethnicity, 2010 and 2050. Retrieved from http://facts.kff.org/chart.aspx?ch=364.
Lurie, N., Harris, K. M., Shih, R. A., Ruder, T., Price, A., Martin, L.G., Acosta, J.C., & Blanchard, J. C. (2010). Assessing Health and Health Care in Prince George's County. Retrieved from www.rand.org
Maryland State Data Center.(2012). 2010 Census Data. Retrieved from http://census.maryland.gov/.
MHCC Extramural Report. (2011). Maryland Physician Workforce Study: Applying the Health Resources and Services Administration Method to Maryland Data. Retrieved from http://mhcc.dhmh.maryland. gov/workforce/Documents/sp.mhcc.maryland.gov/workforce/physician_workforce_study_20110513. pdf
United States Census. (2011). Census 2000 Public Use Microdata Area (PUMA) Maps. Retrieved from http://www.census.gov/geo/www/maps/puma5pct.htm
United States Census Bureau. (2010). American Fact Finder. Retrieved from http://factfinder2.census.gov/ faces/nav/jsf/pages/index.xhtml
Valentine, D. (2008, October 30). County has highest percentage of uninsured residents in Maryland. Gazette.net. Retrieved from http://ww2.gazette.net/stories/10302008/prinnew171042_32470.shtml
Min Qi Wang, PhD
Audrey Jeanne Babkirk, BA
Fang (Alice) Yan, M.D, PhD
Min Qi Wang, PhD, Professor, Department of Behavioral and Community Health, School of Public Health, University of Maryland, 2387 SPH Building, College Park, MD 20742-2611, Tel:301-405-6652, Fax:301-314-9167, email: firstname.lastname@example.org. Audrey Jeanne Babkirk, BA, Master of Public Health Student, Department of Behavioral and Community Health, School of Public Health, University of Maryland, College Park, MD 20742-2611, Email: email@example.com.Fang (Alice) Yan, M.D, PhD, Assistant Professor, Silber School of Public Health, University of Wisconsin at Milwaukee, Milwaukee, WI 53201, Email: yanf@ uwm.edu
Figure 3 Age-Adjusted Myocardial Infarction Rate per 100,000 Residents by Jurisdiction, 2009 Maryland Jurisdiction Anne Arundel 174.1 Baltimore 190.6 Howard 120.7 Montgomery 91.7 Prince 110.0 George's Maryland Average = 81.8 Note: Table made from bar graph. Figure 5. Age-Adjusted Asthma Rate per 100,000 Residents by Jurisdiction, 2009 Maryland Jurisdiction Anne Arundel 137.2 Baltimore 233.1 Howard 74.0 Montgomery 90.1 Prince 127.4 George's Maryland Average = 83.8 Note: Table made from bar graph. Figure 7. Primary Care Physician Rate per 100,000 Residents by Jurisdiction, 2010 Maryland Jurisdiction Anne Arundel 66.6 Baltimore 101.2 Howard 75.2 Montgomery 94.9 Prince 53.9 George's Maryland Average = 84.4 Note: Table made from bar graph.
Please note: Some tables or figures were omitted from this article.
|Printer friendly Cite/link Email Feedback|
|Author:||Wang, Min Qi; Babkirk, Audrey Jeanne; Yan, Fang "Alice"|
|Publication:||American Journal of Health Studies|
|Date:||Mar 22, 2012|
|Previous Article:||Role of health belief model on sexual communication among African immigrants.|
|Next Article:||Sex vs. gender: cultural competence in health education research.|