Show which would be predicted by Wennbergargument Diabetes hospitalizations were significantly predicted by small areas with fewer rural residents and lower high school graduation rates Discussion Wennbergin the Utah which can regression Greenland finding received mixed support data Although the conclusions be made from ecological Rothman are limited 1998 it appears that local socioeconomic status and rurality are independent predictors of these selected ACSH in Utah beyond the effects of local bed supply However the effects of rural residence were to decrease hospitalizations not increase them The use of ACSH as administrative indicators of access to primary care should be used with caution since bed supply has significant effects Where there are many hospital beds there are also many primary care providers so thathigh ACSH rate in these areas does not necessarily reflect low numbers of providers Number of providers is not the only feature of access to primary care and the Utah data show that ACSH are also elevated in areas of low socioeconomic status Lack of insurance is notsignificant factor after SES is adjusted but local education levels seem to be key This may be due to collinearity among these various features of SES or it may implicate an access barrier related to education perhaps one such as ability to recognize symptoms or ability to navigate through the health care system Thus bed supply although related cannot be solely responsible for these hospitalizations Rurality is often considered an indicator of geographic barriers to health care Gesler Ricketts 1992 Others have found that ruralffrontier Utah has ACSH rates which are elevated relative to urban areas Silver Magill Babitz 1997 and our analysis of pneumonia hospitalizations follows this trend although not statistically significantly so The finding of more astima hospitalizations in urban areas has been found in other states Thomas Whitman the finding of higher However 1999 diabetes rates in more urbanized areas is puzzling Although the purpose of the present study is to investigate whether access factors beyond bed supply are related to ACSH in Utah Tableshows evidence of disparity in ACSH by local racelethnicity mix Collinearity with SES in this dataset cannot differentiate access barriers related to race ethnicity mix or SES as determinants of ACSH This suggests that access factors beyond bed supply come into play Additional discussions of race ethnicity and SES can be found in the literature Kaufman Cooper 1999 LaVeist 1994 LillieBlanton LaVeist 1996 There are several limitations of this study One is thatcredible measure of physician supply was not available for these small areas Also asthma and pneumonia hospitalizations were age specific rather than age-adjusted and this may have inadequately controlled for the unequal distribution of the very young or very old cases in small areas In addition the timing of census information and other predictors of hospitalization rates were not entirely synchronous This probably made trends weaker than they would have appeared with complete and concurrent more measurement of features of access Another limitation was that two small areas were excluded from analysis because they were created after the 1990 census These were due to new housing developments as urban areas expanded In spite of these limitations this analysis still indicates that features of access to primary care are correlated with ACSH in small areas of this geographically rurallfrontier state adjusting for bed supply Thus the use of ACSH as an administrative warning flag of access to primary care in the 61 small areas of Utah Haggard Shah Rolfs 1999 was supported by this analysis Digital image 2005 Marriott Library University of Utah Al rights reserved |