Regions in Texas with only RHCs have the highest rates of hospitalization for COPD compared with regions with only FQHCs, which have the lowest rates of hospitalization. Variations in access to health care may partly explain these disparities. Regions with only RHCs are mostly rural and have the lowest access to pulmonary specialists and hospitals. In contrast, counties with only FQHCs are near major metropolitan areas in Texas, and have greater access to pulmonary specialists and hospitals. Interestingly, counties with both types of health centers and counties with neither had similar intermediate hospitalization rates. These health centers are located in areas with medically underserved populations and have intermediate-density levels of hospitals, pulmonary specialists, and rurality. Health centers are placed in medically underserved areas; counties with both types of health centers may have a greater medically underserved population. The presence of the facilities does not necessarily mean that people are utilizing them. Having both types of facilities in a county may be a reflection of a significant need for treating other diseases, not necessarily COPD.
Factors that may affect COPD hospitalization rates are many and include patient-related, environmental, and health care factors. Patient-related factors include: gender, race/ethnicity, severity of illness, low adherence to treatments that decrease exacerbations, and delay in treating exacerbations with systemic steroids and antibiotics. The significant differences between genders observed in counties with only FQHCs or RHCs showed that females had approximately 10% increase in hospitalization rates. Different economic and psychosocial factors may have influenced the utilization patterns of individuals in these areas. Gender differences in manifestations of COPD have suggested that females may experience changes in lung function and susceptibility, while males may experience more cough and sputum.24
Investigation of health care utilization among COPD patients has found that over 50% of emergency department and inpatient admissions are made by females.26
The significantly lower rates observed among Hispanics may be due to their lower prevalence of COPD.27
The RR for Hispanics relative to non-Hispanic White hospitalization is similar across all strata except for counties that have both types of health clinics, which are also predominantly Hispanic (52.72%).
Health care related factors include access to and quality of primary, specialty, and hospital care. The differences observed in rates may have been affected by variations in staffing and services at the health centers. RHCs must be staffed by at least one mid-level practitioner (physician assistant, nurse practitioner, or certified nurse midwife), and a physician must be present to supervise the practitioner. The mid-level staff must be onsite at least 50% of the time the clinic is open.29
While FQHCs have no specific staffing requirements, they are required to have a core staff of fulltime providers, but no definition of core staff. It is recommended that they maintain a staff level that allows for about 5000 visits per year, with a physician to patient ratio of 1:1500 and a mid-level practitioner to patient ratio of 1:750.30
Moreover, differences in services provided between the two types of health clinics may have an effect on hospitalization rates. FQHCs have a governing board that is made up of at least 50% of registered clients, which enables the health center to set agendas for improving the health status of the community. FQHCs are also required to provide information on preventive care. There has been a large investment in FQHCs but not in RHCs.31
In Texas, counties with RHCs have poorer health outcomes overall in terms of mortality and morbidity than the counties of major metropolitan areas.32
The unadjusted RR for counties with a pulmonary specialist showed that there was a significantly lower rate of hospitalization. This was also seen in the stratified analysis for counties with only FQHCs, and counties with both FQHCs and RHCs. These results are consistent with evidence that physician supply and ACS conditions are inversely related for patients 18 years and older.33
The results of this exploratory analysis have some limitations. Smoking status and smoking rates for subgroups/regions were not available; a higher prevalence of smoking would contribute to a higher prevalence of COPD and exacerbations, which in turn leads to increased hospitalization rates. A study of the prevalence of smoking found that the rate was highest in the Southern United States, and among those states, rural Texas smoking rates had increased by 2% between the mid-1990s and 2001.34
In addition to this, the data obtained from the THCIC did not contain unique patient identifiers, which prohibited the analysis from differentiating between repeated hospitalizations. ICD-9 codes were used to identify cases of COPD because spirometry data was not available, which may result in diagnostic misclassification. However, the use of these codes as a proxy for case identification has been used in other studies.7
Moreover, the lack of spirometry data did not allow us to determine the severity of the disease, which affects frequency of exacerbations. The information available did not allow us to determine whether or not a patient had used a health clinic in their county. Knowledge of which health clinic was used might affect the results, since use of a clinic in a contiguous county is probable if there is no clinic nearby. The way in which the medically underserved areas were defined for clinics was not at the county level, but at a smaller scale, which data limitations did not allow us to take into account. The analysis dataset contained a combination of individual and aggregate level variables, interpretations may be subject to ecologic fallacy.