Abstract
Background
Rural residence has been associated with a lower incidence of inflammatory bowel disease (IBD) but higher health care utilization and worse outcomes. Socioeconomic status is intrinsically tied to both IBD incidence and outcomes. Inflammatory bowel disease outcomes have not been investigated in Appalachia: a rural, economically distressed region rife with risk factors for both increased incidence and unfavorable outcomes.
Methods
Hospital inpatient discharge and outpatient services databases were utilized to assess outcomes in patients diagnosed with either Crohn’s disease (CD) or ulcerative colitis (UC) in Kentucky. Encounters were classified by patient residence in Appalachian or non-Appalachian counties. Data were reported as crude and age-adjusted rates of visits per 100,000 population per year collected in 2016 to 2019. National inpatient discharge data from 2019, stratified by rural and urban classification codes, were utilized to compare Kentucky to national trends.
Results
Crude and age-adjusted rates of inpatient, emergency department and outpatient encounters were higher in the Appalachian cohort for all 4 years observed. Appalachian inpatient encounters are more frequently associated with a surgical procedure (Appalachian, 676, 24.7% vs non-Appalachian, 1408, 22.2%; P = .0091). In 2019, the Kentucky Appalachian cohort had significantly higher crude and age-adjusted rates of inpatient discharges for all IBD diagnoses compared with national rural and nonrural populations (crude 55.2; 95% CI, 50.9-59.5; age-adjusted 56.7; 95% CI, 52.1-61.3).
Conclusions
There is disproportionately higher IBD health care utilization in Appalachian Kentucky compared with all cohorts, including the national rural population. There is a need for aggressive investigation into root causes of these disparate outcomes and identification of barriers to appropriate IBD care.
Keywords: inflammatory bowel disease, health care disparities, outcomes and utilization
Key Messages
What is already known? Rural residence is associated with a “protective effect” against IBD, though higher IBD-related health care utilization and worse outcomes have been reported in rural areas.
What is new here? The Appalachian region of Kentucky has higher rates of health care utilization than non-Appalachian Kentucky, as well as the national rural cohort, suggesting rural residence alone is not driving outcomes in this region.
How can this study help patient care? The findings in this study will fuel further investigation into causes for IBD outcome disparity in this region. Furthermore, the methods utilized in this study can be used by others to identify outcome disparity in other regions of the country.
Introduction
Inflammatory bowel disease (IBD), comprising Crohn’s disease (CD) and ulcerative colitis (UC), is a chronic illness that follows a relapsing and remitting course throughout a person’s lifetime. Inflammatory bowel disease is estimated to affect up to 3.1 million US adults.1–3 Greater disease prevalence is observed in subpopulations of non-Hispanic white adults, as well as patients with lower educational attainment, unemployment status, and lower socioeconomic status.2,4,5 Prompt disease recognition, regular monitoring, and tight control of disease activity are required to reduce disease morbidity and improve outcomes, especially in vulnerable subpopulations.6 Poor disease control results in adverse outcomes, including increased rates of surgical intervention, opioid and steroid prescription use, as well as hospitalizations and emergency department (ED) visits.7 Severe CD is also associated with a distinct increase in mortality.8
Appalachia is a low-resourced area with a geographically isolated population that is uniquely at risk for poor health outcomes.9 The Appalachian Region is a “205,000-sqare-mile region that follows the spine of the Appalachian Mountains from southern New York to northern Mississippi” and is home to over 25 million people.10 The Appalachian Regional Commission, a federal-state government partnership aimed at improving conditions in the region, further categorizes these counties into 5 groups based on their economic development ranging from distressed counties that rank in the lowest 10% of the nation’s counties to attainment counties ranking in the nation’s top 10%. Most of Kentucky’s Appalachian counties are considered distressed, which is associated with overall poorer health outcomes.11,12
Social determinants of health are correlated with IBD prognosis. The demographic and socioeconomic composition of Appalachia puts this subpopulation at unique risk for increased prevalence of IBD and morbidity from their disease.4 Appalachia’s population is composed of proportionally more non-Hispanic whites, and the 4-year college completion rate is lower compared with the US average.13 Though not universally true of broader Appalachia, the Kentucky Appalachian region is considered rural. Although rural residence in early life has been associated with lower IBD incidence and a possible “protective effect,” there remain risk factors for worse outcomes in rural patients.14,15 Rural residence is associated with decreased access to specialty health care providers, including gastroenterologists.16 Health care facilities in the region are limited to primary care and critical-access hospitals, which are less likely to have access to IBD expertise. Lack of infrastructure and reliable transportation limit access to IBD centers.13 Furthermore, Appalachian Kentuckians are less likely to have diets high in fruits and vegetables and are more likely to consume foods that are high in fat.17,18 These high-fat, low-fiber diets are associated with increased risk of IBD.19–21 Subsequently, despite the potential protective effect of the rural locale, there are risk factors for increased IBD prevalence in the Appalachian region. Limited access to health care resources and IBD expertise increases risk for poorer disease outcomes in this population.
Despite a higher prevalence of risk factors for poor IBD outcomes in Appalachian Kentucky, an investigation of disparities in outcomes for this region has not been performed. There is a paucity of data evaluating IBD outcomes in rural populations, though a few studies have suggested rural residence may have a protective effect.14,22 Despite this, others have noted increased health care utilization among IBD patients residing in rural areas. In the United States, a recent study assessing the impact of rural vs urban residence on health care utilization for patients with IBD found that CD patients residing in rural areas were more likely than their urban counterparts to visit the ED or be hospitalized for their disease—but less likely to see an outpatient gastroenterologist.23 Similarly, Benchimol and colleagues directly compared health care utilization and associated costs between rural and urban IBD populations. This study reported higher rates of health care utilization and higher direct costs in the rural IBD cohort.24,25 In consideration of the conflicting evidence, further investigation into rural disparities in outcomes for IBD is warranted.
This study aims to identify disparities in health care outcomes and utilization between Appalachian and non-Appalachian Kentuckians with IBD. We hypothesize patients residing in Appalachian Kentucky have worse outcomes compared with both non-Appalachian Kentuckian and national populations, as measured by increased rates of inpatient admissions, ED visits, surgeries, and IBD-related complications. To investigate this hypothesis, we designed a study to examine IBD outcomes in Kentucky stratified by Appalachian residence. Additionally, IBD outcomes were evaluated in a national sample, stratifying by rural vs nonrural place of residence. We hypothesize Appalachian Kentuckians with IBD have worse outcomes and higher rates of health care utilization than non-Appalachian Kentuckians. Furthermore, we posit the disparity in IBD outcomes detected in Appalachian counties is greater than reported in rural patients nationally.
Materials and Methods
Data Sources
This study utilized the Kentucky Hospital Inpatient Discharge and Outpatient Services Databases (HIDOSD) from 2016 to 2019 provided by the Kentucky Cabinet for Health and Family Services.26,27 These databases contain de-identified encounter-level inpatient data and ED discharges in Kentucky, as well as services received at ambulatory facilities. This data set represents greater than 99% of health care transactions described in Kentucky and is analogous to state-level databases maintained by the Healthcare Cost and Utilization Project (HCUP). In addition to Kentucky HIDOSD, national-level data were collected from the HCUP National Inpatient Sample (NIS) for 2019. The NIS contains weighted encounter-level data representative of greater than 35 million hospitalizations per year nationwide in the United States.28 Data collection and analysis for this study were performed under a research repository IRB protocol for the respective databases at our institution.
Study Design and Summary
This study is a retrospective, multicohort design classifying encounter data into Kentucky Appalachian, Kentucky non-Appalachian, and national cohorts. The national cohorts were further stratified by rural vs nonrural. These cohorts were compared to accomplish the following aims: (1) evaluate disparities in health care utilization and clinical outcomes for IBD between Kentucky Appalachian and non-Appalachian residents; and (2) contextualize the health care utilization and clinical outcomes for IBD in Kentucky compared with national trends.
To accomplish the first aim, we analyzed Kentucky inpatient and outpatient encounters utilizing Kentucky HIDOSD data. We identified health care encounters of interest (inpatient admissions, ED, and outpatient clinic visits) for patients with IBD from 2016 through 2019 from HIDOSD database. These encounters were classified by county of patient residence. Appalachian counties are defined by the Appalachian Regional Commission.13 The primary outcome was the crude and age-adjusted rate of inpatient admissions per 100,000 persons in each population. Secondary outcomes included crude and age-adjusted rate of ED visits and outpatient clinic visits. Tertiary outcomes included diagnoses and procedures performed during inpatient encounters relevant to the disease state. Results were reported in aggregate and subgrouped by disease state (ie, Crohn’s disease and ulcerative colitis).
For the second aim, we performed a secondary analysis utilizing NIS and HIDOSD data from 2019. Due to cost limitations, we limited evaluation of national data to a single year. Inpatient encounters of interest for patients with IBD in 2019 were identified from both databases. National Inpatient Sample encounters were classified by patient residence in a rural or nonrural area as designated by National Center for Health Statistics (NCHS) 2013 Urban-Rural Classification codes (NCHS 2013 >4 designating rural), whereas HIDOSD data was classified by Appalachian residence as described in the primary analysis.29,30 The primary outcome was the crude and age-adjusted rates of inpatient admissions per 100,000 persons in each cohort. Secondary outcomes included diagnoses and procedures performed during inpatient encounters relevant to the disease state. Results were reported in aggregate and subgrouped by disease state (ie, Crohn’s disease and ulcerative colitis).
Data Collection and Study Definitions
We applied inclusion and exclusion criteria to identify encounters of interest. For the primary analysis, inpatient encounters of interest were identified with discharge dates between January 1, 2016, and December 31, 2019, with a Medicare Severity Diagnosis Related Group (MSDRG) related to gastrointestinal health, and at least 1 diagnosis code for Crohn’s disease (ICD-10 K50x) or ulcerative colitis (ICD-10 K51x; Supplemental Table 1). Outpatient encounters of interest were identified as possessing at least 1 diagnosis code for Crohn’s disease or ulcerative colitis utilizing the same definitions. Encounters were excluded if they were missing relevant data to classify patient residence as Appalachian/non-Appalachian (ie, ZIP/FIPS code) or indicated patient residence outside of Kentucky.
In the secondary analysis, inpatient encounters of interest were identified from both the HIDOSD and NIS as having a discharge date between January 1, 2019, and December 31, 2019, utilizing the same MSDRG and diagnosis code criteria as the primary analysis. Encounters were excluded for HIDOSD encounters as described in the primary analysis. Encounters from the NIS were only excluded if they were missing NCHS 2013 codes in order to classify patient residence as rural/nonrural.
Procedures of interest were identified by use of HCUP Clinical Classification Software (CCS), which classifies ICD-10-PCS codes into clinically meaningful categories,31 and diagnoses of interest were identified utilizing ICD-10-CM codes. Clinical Classification Software category definitions utilized to identify gastrointestinal surgeries and endoscopies were selected by the investigators using this tool. The MSDRG, ICD-10-CM codes, and CCS procedure definitions used in this study are included in Supplemental Table 1.
Statistical Methodology
Rates of health care utilization were calculated similarly to previous methods promulgated by HCUP.32 Incidence rates per 100,000 population per year were obtained by utilizing estimates of county population within each cohort from the US Census Bureau for years 2016 to 2019 as the denominator.33 Age-adjusted rate analyses were conducted by direct standardization to the 2000 US Standard Population.34 Age groups selected to be used in the age-adjustment were defined to best fit existing HIDOSD age groupings, as HIDOSD does not provide exact patient age. Both crude and age-adjusted rates are compared for statistical significance by determining variances of rate estimates and subsequently constructing 95% confidence intervals.35,36 Heat maps of crude rates of health care utilization by county were constructed using Tableau 2021.4 (Mountain View, California).
Diagnoses and procedures occurring during inpatient admissions were collected as binary variables indicating the presence of at least 1 specified procedure or diagnosis code of the given category during the admission. Diagnoses and procedures are compared utilizing Pearson’s χ2 test to identify any differences in patient presentation or services offered, utilizing an alpha level of 0.05 for significance. All analyses were conducted in SAS 9.4 (Cary, North Carolina). Age-adjusted rate analyses were performed utilizing PROC SURVEYMEANS SAS procedures.
Results
Kentucky Health Care Utilization Rate Analysis
Crude and age-adjusted rates of health care utilization in Kentucky for years 2016 to 2019 are presented in Table 1. Significantly higher crude rates of inpatient admissions (Appalachian 58.8; 95% CI, 56.5-61; non-Appalachian 48.1; 95% CI, 46.9-49.3), ED visits (Appalachian 177.8; 95% CI, 173.9-181.7; non-Appalachian 135.8; 95% CI, 133.8-137.8), and outpatient encounters (Appalachian 868.9; 95% CI, 860.4-877.5; non-Appalachian 681.5; 95% CI, 677-686) for IBD per 100,000 population per year were observed across all 4 years. Crude rate analyses for health care utilization for each year examined are displayed in Figure 1. When subgrouped by disease state, the rate of inpatient, ED, and outpatient encounters for CD is also significantly higher for all 4 years. For UC, although there is no significant difference in rate of inpatient encounters, there is a significantly higher rate of ED and outpatient encounters.
Table 1.
Rates of health care utilization for inflammatory bowel disease for all years in Kentucky (2016-2019) by Appalachian residence.
| Primary Diagnosis | Utilization Type | Geographic Cohort | Number of Discharges (2016-2019) |
Crude rate | Standard Error |
95% CI | Age-adjusted rate | Standard Error |
95% CI |
|---|---|---|---|---|---|---|---|---|---|
| Any IBD | |||||||||
| Inpatient Admissions | non-Appalachian | 6350 | 48.1 | 0.60 | (46.9-49.3) | 51.7 | 0.67 | (50.3-53) | |
| Appalachian | 2739 | 58.8 | 1.12 | (56.5-61) | 62.1 | 1.23 | (59.7-64.6) | ||
| ED Visits | non-Appalachian | 17919 | 135.8 | 1.01 | (133.8-137.8) | 150.5 | 1.16 | (148.2-152.8) | |
| Appalachian | 8286 | 177.8 | 1.95 | (173.9-181.7) | 192.7 | 2.19 | (188.4-197) | ||
| Outpatient Visits | non-Appalachian | 89891 | 681.5 | 2.27 | (677-686) | 717.7 | 2.48 | (712.8-722.6) | |
| Appalachian | 40496 | 868.9 | 4.31 | (860.4-877.5) | 911.0 | 4.69 | (901.7-920.3) | ||
| Crohn’s Disease | |||||||||
| Inpatient Admissions | non-Appalachian | 4357 | 33.0 | 0.50 | (32-34) | 35.8 | 0.56 | (34.7-36.9) | |
| Appalachian | 1958 | 42.0 | 0.95 | (40.1-43.9) | 44.4 | 1.04 | (42.4-46.5) | ||
| ED Visits | non-Appalachian | 14953 | 113.4 | 0.93 | (111.5-115.2) | 125.8 | 1.06 | (123.7-127.9) | |
| Appalachian | 6924 | 148.6 | 1.78 | (145-152.1) | 160.6 | 1.99 | (156.7-164.5) | ||
| Outpatient Visits | non-Appalachian | 58835 | 446.0 | 1.84 | (442.4-449.7) | 472.6 | 2.02 | (468.6-476.6) | |
| Appalachian | 28695 | 615.7 | 3.63 | (608.5-622.9) | 647.5 | 3.96 | (639.6-655.3) | ||
| Ulcerative Colitis | |||||||||
| Inpatient Admissions | non-Appalachian | 2042 | 15.5 | 0.34 | (14.8-16.2) | 16.3 | 0.37 | (15.6-17) | |
| Appalachian | 799 | 17.1 | 0.61 | (15.9-18.3) | 18.1 | 0.67 | (16.8-19.5) | ||
| ED Visits | non-Appalachian | 3029 | 23.0 | 0.42 | (22.1-23.8) | 25.2 | 0.48 | (24.3-26.2) | |
| Appalachian | 1392 | 29.9 | 0.80 | (28.3-31.5) | 32.8 | 0.92 | (31-34.6) | ||
| Outpatient Visits | non-Appalachian | 31380 | 237.9 | 1.34 | (235.2-240.5) | 247.7 | 1.45 | (188.4-197) | |
| Appalachian | 11962 | 256.7 | 2.35 | (252-261.3) | 267.1 | 2.54 | (188.4-197) | ||
Figure 1.
Crude rates of inpatient discharges, ED discharges, and outpatient visits between Appalachian and non-Appalachian Kentucky cohorts (2016-2019).
Results of the age-adjusted rate analysis are comparable with the crude analysis, with significantly higher rates of inpatient admissions (Appalachian 62.1, 95% CI, 59.7-64.6; non-Appalachian 51.7, 95% CI, 50.3 -53), ED visits (Appalachian 192.7, 95% CI188.4-197; non-Appalachian 150.5, 95% CI, 148.2-152.8), and outpatient encounters (Appalachian 911, 95% CI, 901.7-920.3; non-Appalachian 717.7, 95% CI, 712.8-722.6) for IBD per 100,000 population per year. Age-adjusted rate analyses for each examined year are displayed in Figure 2. The trends in crude and age-adjusted rate analyses year-on-year suggest a decreasing rate of inpatient discharges and an increasing rate of ED and outpatient visits over the course of the 4 years evaluated.
Figure 2.
Age-adjusted rates of inpatient discharges, ED discharges, and outpatient visits between Appalachian and non-Appalachian Kentucky cohorts (2016-2019).
To geographically delineate areas of increased IBD-related health care utilization, heat maps displaying crude rates of inpatient, ED, and outpatient encounters at the county level were created (Figure 3).
Figure 3.
Heat maps illustrating prevalence of inpatient discharges, ED discharges, and outpatient visits per 100,000 persons per year by county (2016-2019). The counties that comprise the Appalachian portion of the state are outlined in black.
Kentucky Inpatient Encounter Diagnoses and Procedures
A total of 9089 inpatient discharges for IBD were identified between 2016 and 2019, with 2739 (30.1%) of those discharges associated with a patient residence in an Appalachian county (Table 2). Crohn’s disease was the predominant IBD encounter diagnosis (6315 discharges, 69.5%). Appalachians with IBD were more likely than their non-Appalachian counterparts to be white (P < .001), insured by Medicaid/Medicare (P < .0001), and to have a discharge diagnosis of CD (P = .0064).
Table 2.
Demographics of Kentucky inpatient discharges for inflammatory bowel disease, 2016-2019.
| Demographics | |||||
|---|---|---|---|---|---|
| Attribute | Appalachian (n = 2739) |
non-Appalachian (n = 6350) |
Overall (n = 9089) |
P | |
| Inflammatory Bowel Disease | |||||
| Crohn’s disease | 1958 (71.5%) | 4357 (68.6%) | 6315 (69.5%) | 0.0064* | |
| Ulcerative colitis | 799 (29.2%) | 2042 (32.2%) | 2841 (31.3%) | 0.0048* | |
| Gender | |||||
| Female | 1545 (56.4%) | 3742 (58.9%) | 5287 (58.2%) | 0.0253* | |
| Race | |||||
| White | 2661 (97.2%) | 5352 (84.3%) | 8013 (88.2%) | ||
| Black | 43 (1.6%) | 849 (13.4%) | 892 (9.8%) | ||
| Hispanic | 18 (0.7%) | 78 (1.2%) | 96 (1.1%) | <0.0001* | |
| Asian or Pacific Islander | 2 (0.1%) | 36 (0.6%) | 38 (0.4%) | ||
| Native American | 0 (0%) | 4 (0.1%) | 4 (<0.01%) | ||
| Other | 15 (0.6%) | 31 (0.5%) | 46 (0.5%) | ||
| Payor | |||||
| Medicaid | 812 (29.7%) | 1531 (24.1%) | 2343 (25.8%) | <0.0001* | |
| Medicare | 1021 (37.3%) | 1992 (31.4%) | 3013 (33.2%) | <0.0001* | |
| Commercial | 800 (29.2%) | 2579 (40.6%) | 3379 (37.2%) | <0.0001* | |
| Other | 106 (3.9%) | 248 (3.9%) | 354 (3.9%) | 0.936 | |
| Inflammatory Bowel Disease Risk Factors | |||||
| Attribute |
Appalachian
(n = 2739) |
Non-Appalachian
(n = 6350) |
Overall
(n = 9089) |
P | |
| Tobacco Use | |||||
| Overall | 727 (26.5%) | 1481 (23.3%) | 2208 (24.3%) | 0.001* | |
| Crohn’s disease | 608 (31.1%) | 1182 (27.1%) | 1790 (28.4%) | 0.0014* | |
| Ulcerative colitis | 124 (15.5%) | 309 (15.1%) | 433 (15.2%) | 0.7963 | |
| Obesity | |||||
| Overall | 303 (11.1%) | 638 (10.1%) | 941 (10.4%) | 0.1449 | |
| Crohn’s disease | 198 (10.1%) | 413 (9.5%) | 611 (9.7%) | 0.431 | |
| Ulcerative colitis | 106 (13.3%) | 233 (11.4%) | 339 (11.9%) | 0.17 | |
| Opioid Use Disorder | |||||
| Overall | 83 (3%) | 109 (1.7%) | 192 (2.1%) | <0.0001* | |
| Crohn’s disease | 70 (3.6%) | 92 (2.1%) | 162 (2.6%) | 0.0007* | |
| Ulcerative colitis | 14 (1.8%) | 18 (0.9%) | 31 (1.1%) | 0.0480* | |
| Inflammatory Bowel Disease-Related Diagnoses | |||||
| Attribute |
Appalachian
(n = 2739) |
Non-Appalachian
(n = 6350) |
Overall
(n = 9089) |
P | |
| Intestinal abscess | |||||
| Overall | 127 (4.6%) | 343 (5.4%) | 470 (5.2%) | 0.1308 | |
| Crohn’s disease | 103 (5.3%) | 273 (6.3%) | 376 (6%) | 0.1184 | |
| Ulcerative colitis | 25 (3.1%) | 72 (3.5%) | 97 (3.4%) | 0.6003 | |
| Anal abscess | |||||
| Overall | 45 (1.6%) | 109 (1.7%) | 154 (1.7%) | 0.803 | |
| Crohn’s disease | 40 (2%) | 93 (2.1%) | 133 (2.1%) | 0.8146 | |
| Ulcerative colitis | 6 (0.8%) | 18 (0.9% | 24 (0.8%) | 0.7325 | |
| Intestinal fistula | |||||
| Overall | 22 (0.8%) | 77 (1.2%) | 99 (1.1%) | 0.0845 | |
| Crohn’s disease | 20 (1%) | 73 (1.7%) | 93 (1.5%) | 0.046* | |
| Ulcerative colitis | 2 (0.3%) | 4 (0.2%) | 6 (0.2%) | 0.7763 | |
| Intestinal perforation | |||||
| Overall | 30 (1.1%) | 85 (1.3%) | 115 (1.3%) | 0.341 | |
| Crohn’s disease | 26 (1.3%) | 67 (1.5%) | 93 (1.5%) | 0.5219 | |
| Ulcerative colitis | 4 (0.5%) | 18 (0.9%) | 22 (0.8%) | 0.2978 | |
| Abdominal pain | |||||
| Overall | 128 (4.7%) | 276 (4.4%) | 404 (4.4%) | 0.4879 | |
| Crohn’s disease | 99 (5.1%) | 207 (4.8%) | 306 (4.9%) | 0.6014 | |
| Ulcerative colitis | 32 (4%) | 73 (3.6%) | 105 (3.7%) | 0.5849 | |
| Intestinal obstruction | |||||
| Overall | 245 (8.9%) | 591 (9.3%) | 836 (9.2%) | 0.5835 | |
| Crohn’s disease | 201 (10.3%) | 494 (11.3%) | 695 (11%) | 0.2078 | |
| Ulcerative colitis | 44 (5.5%) | 99 (4.9%) | 143 (5%) | 0.4703 | |
| Postoperative malabsorption | |||||
| Overall | 37 (1.4%) | 112 (1.8%) | 149 (1.6%) | 0.1549 | |
| Crohn’s disease | 31 (1.6%) | 95 (2.2%) | 126 (2%) | 0.1165 | |
| Ulcerative colitis | 6 (0.8%) | 17 (0.8%) | 23 (0.8%) | 0.8273 | |
| Mortality | |||||
| Overall | 14 (0.5%) | 29 (0.5%) | 43 (0.5%) | 0.7285 | |
| Crohn’s disease | 6 (0.3%) | 13 (0.3%) | 19 (0.3%) | 0.9568 | |
| Ulcerative colitis | 8 (1%) | 16 (0.8%) | 24 (0.8%) | 0.5686 | |
| Inflammatory Bowel Disease-Related Procedures | |||||
| Attribute |
Appalachian
(n = 2739) |
Non-Appalachian
(n = 6350) |
Overall
(n = 9089) |
P | |
| Surgery | |||||
| Overall | 676 (24.7%) | 1408 (22.2%) | 2084 (22.9%) | 0.0091* | |
| Crohn’s disease | 497 (25.4%) | 1083 (24.9%) | 1580 (25%) | 0.655 | |
| Ulcerative colitis | 180 (22.5%) | 331 (16.2%) | 511 (18%) | <0.0001* | |
| Endoscopy | |||||
| Overall | 661 (24.1%) | 1633 (25.7%) | 2294 (25.2%) | 0.1108 | |
| Crohn’s disease | 374 (19.1%) | 888 (20.4%) | 1262 (20%) | 0.2394 | |
| Ulcerative colitis | 299 (37.4%) | 768 (37.6%) | 1067 (37.6%) | 0.9257 | |
*Denotes statistical significance.
Appalachian inpatient encounters were more frequently associated with diagnoses of IBD risk factors of tobacco use (Appalachian discharges, 727, 26.5%; vs non-Appalachian discharges, 1481, 23.3%; P = .001) and opioid use disorder (Appalachian discharges, 83, 3%; vs non-Appalachian discharges, 109, 1.7%; P < .0001).
There were no significant differences in IBD-related complications (abscess, perforation, obstruction, fistualization, malabsorption, abdominal pain) in the general study population (Table 2). This remained true when analyzing complications by disease subgroup apart from intestinal fistula, where fewer Appalachian encounters were associated with an intestinal fistula diagnosis compared with non-Appalachian patients (Appalachian discharges, 20, 1%; vs non-Appalachian discharges, 73, 1.7%; P = .046).
Appalachian inpatient encounters were more commonly associated with a surgical procedure (Appalachian discharges , 676, 24.7%; vs non-Appalachian discharges, 1408, 22.2%; P = .0091). This difference was driven by inpatient encounters for UC, with 22.5% of Appalachians with UC requiring a surgical procedure vs 16.2% of non-Appalachian UC patients during the study period (P < .0001). There were no differences in the proportion of encounters with inpatient endoscopic procedures performed between each cohort. No significant difference was observed in inpatient mortality rates (P = .7285).
Comparison of Kentucky Inpatient Utilization Rate to National Trends
In the crude rate analysis of year 2019, the Kentucky Appalachian cohort has a significantly higher rate of inpatient discharges for all IBD diagnoses compared with the Kentucky non-Appalachian, as well as national rural and nonrural populations (55.2 discharges per 100,000 population; 95% CI, 50.9-59.5; Table 3). This was driven by the rate of inpatient discharges associated with CD in the Kentucky Appalachian cohort, which was significantly higher than other cohorts. No significant difference was observed between cohorts in rate of inpatient discharges associated with UC. No significant difference in the rate of health care utilization was observed between national rural and nonrural cohorts for IBD or in the subgroup of encounters for CD and UC.
Table 3.
Rate of discharges for inflammatory bowel disease in 2019 by geographic location
| Primary Diagnosis | Geographic Cohort | Number of Discharges in 2019 |
Crude rate | Standard Error |
95% CI | Age-adjusted rate | Standard Error |
95% CI |
|---|---|---|---|---|---|---|---|---|
| Any IBD | ||||||||
| National nonrural (NCHS Rural-Urban ≥ 4) | 130710 | 46.3 | 0.95 | (44.4-48.2) | 45.0 | 0.94 | (43.1-46.8) | |
| National rural (NCHS Rural-Urban ≤ 5) | 21180 | 46.0 | 1.65 | (42.7-49.2) | 43.7 | 1.61 | (40.5-46.9) | |
| Kentucky non-Appalachian | 1535 | 46.4 | 1.18 | (44-48.7) | 49.7 | 1.31 | (47.1-52.3) | |
| Kentucky Appalachian | 641 | 55.2 | 2.18 | (50.9-59.5) | 56.7 | 2.32 | (52.1-61.3) | |
| Crohn’s Disease | ||||||||
| National nonrural (NCHS Rural-Urban ≥ 4) | 80660 | 28.6 | 0.61 | (27.4-29.8) | 28.2 | 0.62 | (27-29.4) | |
| National rural (NCHS Rural-Urban ≤ 5) | 13790 | 29.9 | 1.16 | (27.6-32.2) | 28.9 | 1.14 | (26.6-31.1) | |
| Kentucky non-Appalachian | 1036 | 31.3 | 0.97 | (29.4-33.2) | 34.0 | 1.09 | (31.8-36.1) | |
| Kentucky Appalachian | 431 | 37.1 | 1.79 | (33.6-40.7) | 38.2 | 1.09 | (36-40.3) | |
| Ulcerative Colitis | ||||||||
| National nonrural (NCHS Rural-Urban ≥ 4) | 50655 | 17.9 | 0.39 | (17.2-18.7) | 17.0 | 0.39 | (16.2-17.8) | |
| National rural (NCHS Rural-Urban ≤ 5) | 7505 | 16.3 | 0.66 | (15-17.6) | 15.1 | 0.63 | (13.8-16.3) | |
| Kentucky non-Appalachian | 503 | 15.2 | 0.68 | (13.9-16.5) | 15.9 | 0.73 | (14.4-17.3) | |
| Kentucky Appalachian | 216 | 18.6 | 1.27 | (16.1-21.1) | 19.1 | 1.34 | (16.4-21.8) | |
In the age-adjusted analysis in 2019, the Kentucky Appalachian and non-Appalachian rates of inpatient discharges for IBD were not significantly different (Kentucky Appalachian 56.7, 95% CI, 52.1-61.3; vs Kentucky non-Appalachian 49.7, 95% CI, 47.1-52.3). However, age-adjusted rates of inpatient discharges for both Kentucky cohorts were significantly higher than national rural and nonrural populations. This trend was observed in the subgroup of discharges for CD but not UC.
Comparison of Kentucky and National Inpatient Encounter Diagnoses and Procedures
A total of 151,890 weighted inpatient discharges for IBD were estimated nationally in 2019 from the NIS; 2176 inpatient discharges for IBD were identified from the Kentucky HIDOSD database for comparison. Crohn’s disease was the predominant IBD diagnosis in both cohorts; however, there was a significantly larger proportion of CD diagnoses in the Kentucky cohort (Kentucky, 1467, 67.4%; vs national, 94,450, 62.2%; P < .0001). Interestingly, the data revealed a significant difference in gender, racial, and health care payor composition between cohorts, with a larger portion of the Kentucky cohort being female, white, and having government payors compared with the national cohort (Table 4).
Table 4.
Demographics of national vs Kentucky inpatient discharges for inflammatory bowel disease, 2019
| Demographics | ||||
|---|---|---|---|---|
| Attribute | Kentucky (n = 2176) |
National (n = 151,890) |
P | |
| Inflammatory Bowel Disease | ||||
| Crohn’s disease | 1467 (67.4%) | 94450 (62.2%) | <0.0001* | |
| Ulcerative colitis | 719 (33%) | 58160 (38.3%) | <0.0001* | |
| Gender | ||||
| Female | 1266 (58.2%) | 83130 (54.7%) | 0.0013* | |
| Race | ||||
| White | 1912 (87.9%) | 111740 (73.6%) | ||
| Black | 220 (10.1%) | 18795 (12.4%) | ||
| Hispanic | 21 (1%) | 11515 (7.6%) | <0.0001* | |
| Asian or Pacific Islander | 9 (0.4%) | 2205 (1.5%) | ||
| Native American | 0 (0%) | 470 (0.3%) | ||
| Other | 14 (0.6%) | 7165 (4.7%) | ||
| Payor | ||||
| Medicaid | 521 (23.9%) | 26170 (17.2%) | <0.0001* | |
| Medicare | 758 (34.8%) | 46795 (30.8%) | <0.0001* | |
| Commercial | 804 (37%) | 66535 (43.8%) | <0.0001* | |
| Other | 93 (4.3%) | 12230 (8.1%) | <0.0001* | |
| Inflammatory Bowel Disease Risk Factors | ||||
| Attribute | Kentucky (n = 2176) |
National (n = 30378) |
P | |
| Tobacco Use | ||||
| Overall | 517 (23.8%) | 25870 (17%) | <0.0001* | |
| Crohn’s disease | 410 (28%) | 19455 (20.6%) | <0.0001* | |
| Ulcerative colitis | 107 (14.9%) | 6500 (11.2%) | 0.0018* | |
| Obesity | ||||
| Overall | 290 (13.3%) | 15995 (10.5%) | <0.0001* | |
| Crohn’s disease | 181 (12.3%) | 9830 (10.4%) | 0.0164 | |
| Ulcerative colitis | 109 (15.2%) | 6235 (10.7%) | 0.0001* | |
| Opioid Use Disorder | ||||
| Overall | 66 (3%) | 3870 (2.6%) | 0.1544 | |
| Crohn’s disease | 54 (3.7%) | 2990 (3.2%) | 0.2639 | |
| Ulcerative colitis | 13 (1.8%) | 900 (1.6%) | 0.574 | |
| Inflammatory Bowel Disease-Related Diagnoses | ||||
| Attribute |
Kentucky
(n = 2176) |
National
(n = 30378) |
P | |
| Intestinal abscess | ||||
| Overall | 123 (5.7%) | 8235 (5.4%) | 0.6368 | |
| Crohn’s disease | 103 (7%) | 5950 (6.3%) | 0.2594 | |
| Ulcerative colitis | 22 (3.1%) | 2365 (4.1%) | 0.1738 | |
| Anal abscess | ||||
| Overall | 36 (1.7%) | 3055 (2%) | 0.2384 | |
| Crohn’s disease | 27 (1.8%) | 2590 (2.7%) | 0.0354* | |
| Ulcerative colitis | 9 (1.3%) | 470 (0.8%) | 0.1881 | |
| Intestinal fistula | ||||
| Overall | 20 (0.9%) | 1225 (0.8%) | 0.5602 | |
| Crohn’s disease | 18 (1.2%) | 1060 (1.1%) | 0.7058 | |
| Ulcerative colitis | 2 (0.3%) | 180 (0.3%) | 0.8805 | |
| Intestinal perforation | ||||
| Overall | 36 (1.7%) | 1950 (1.3%) | 0.1281 | |
| Crohn’s disease | 29 (2%) | 1390 (1.5%) | 0.1118 | |
| Ulcerative colitis | 7 (1%) | 565 (1%) | 0.9954 | |
| Abdominal pain | ||||
| Overall | 90 (4.1%) | 5370 (3.5%) | 0.1324 | |
| Crohn’s disease | 66 (4.5%) | 3905 (4.1%) | 0.4868 | |
| Ulcerative colitis | 25 (3.5%) | 1500 (2.6%) | 0.1319 | |
| Intestinal obstruction | ||||
| Overall | 201 (9.2%) | 13430 (8.8%) | 0.5192 | |
| Crohn’s disease | 162 (11%) | 10215 (10.8%) | 0.7805 | |
| Ulcerative colitis | 39 (5.4%) | 3270 (5.6%) | 0.8186 | |
| Postoperative malabsorption | ||||
| Overall | 40 (1.8%) | 2225 (1.5%) | 0.1508 | |
| Crohn’s disease | 34 (2.3%) | 1850 (2%) | 0.3255 | |
| Ulcerative colitis | 6 (0.8%) | 380 (0.7%) | 0.5498 | |
| Mortality | ||||
| Overall | 12 (0.6%) | 730 (0.5%) | 0.6465 | |
| Crohn’s disease | 5 (0.3%) | 365 (0.4%) | 0.7850 | |
| Ulcerative colitis | 7 (1%) | 370 (0.6%) | 0.2771 | |
| Inflammatory Bowel Disease-Related Procedures | ||||
| Attribute |
Kentucky
(n = 2176) |
National
(n = 30378) |
P | |
| Surgery | ||||
| Overall | 504 (23.2%) | 35460 (23.4%) | 0.8403 | |
| Crohn’s disease | 379 (25.8%) | 23700 (25.1%) | 0.5152 | |
| Ulcerative colitis | 128 (17.8%) | 11900 (20.5%) | 0.0789 | |
| Endoscopy | ||||
| Overall | 567 (26.1%) | 42540 (28%) | 0.0442* | |
| Crohn’s disease | 302 (20.6%) | 21045 (22.3%) | 0.1214 | |
| Ulcerative colitis | 270 (37.6%) | 21825 (37.5%) | 0.9884 | |
*Denotes statistical significance.
Kentucky inpatient encounters were more frequently associated with diagnoses of IBD risk factors of tobacco use (Kentucky, 517, 23.8%; vs national, 25,870, 17%; P < .0001) and obesity (Kentucky, 290, 13.3%; vs national, 15,995, 10.5%; P < .0001). There was no significant difference in diagnosis of opioid use disorder between the Kentucky and national cohorts (P = .154). There were no significant differences in IBD-related complications in the general study population; however, in the subgroup of encounters for CD, significantly fewer Kentucky encounters were complicated by an anal abscess compared with national patients (Kentucky, 27, 1.8%; vs national, 2590, 2.7%; P = .0354). National inpatient encounters were more frequently associated with an endoscopic procedure (Kentucky, 567, 26.1%; vs national, 42,540, 28%;P = .0442). There was no difference in rates of IBD-related surgical procedures between the Kentucky and national cohorts (P = .804). No significant difference was observed in inpatient mortality rates (P = .6465).
Discussion
This study addresses the hypothesis that patients with IBD residing in Appalachian Kentucky have increased rates of health care utilization and poorer outcomes compared with non-Appalachian Kentuckian and national cohorts. The data presented here indicate that Appalachian Kentucky, an economically distressed region in a rural locale, has higher health care utilization and poorer outcomes than would be expected based on rural residence alone. In the 4 years of state-level data analyzed, we found Appalachian Kentuckians with IBD had higher population and age-adjusted rates of inpatient discharges, ED, and outpatient visits than the non-Appalachian cohort. This difference was driven largely by patients with Crohn’s disease, as health care utilization rates for UC were not statistically different between Appalachian and non-Appalachian Kentuckians. When comparing state-level inpatient discharge data from 2019 to the national discharge data sample, the state of Kentucky had higher rates of IBD-related health care encounters than the national cohort. This difference remained when delineating national rural and nonrural cohorts, suggesting rural residence is not the most significant determinate of hospitalization in the Kentucky Appalachian IBD cohort. Furthermore, we cannot rule out that the higher number of inpatient and ER encounters in the Kentucky Appalachian cohort may reflect a higher prevalence of IBD in this region, which would challenge the historic notion of a “rural protective effect.”
Although there was no significant difference in the proportion of encounters with surgical procedures performed in the overall IBD cohorts (both Appalachian vs Non-Appalachian Kentucky as well as Kentucky vs national), Appalachian Kentuckians with ulcerative colitis did have a significantly higher proportion of encounters with surgical interventions than the non-Appalachian cohort. Whether this is due to more severe, refractory disease at admission requiring urgent colectomy or if this reflects inadequate utilization of biologic therapies will require further investigation. Potentially, patients with UC in Appalachian Kentucky are more apt to choose definitive therapy for their disease to avoid difficulties coordinating frequent doctor visits, infusion appointments, and financial barriers.
There were no meaningful differences in the proportion of encounters with IBD-related complications (abscess, perforation, obstruction, etc.) in the Appalachian vs non-Appalachian cohorts, apart from a slight increase in rates of intestinal fistualization in the Non-Appalachian Kentucky cohort. However, significant differences in the proportion of IBD-related complications were also not appreciated when comparing the Kentucky cohorts to the national data sample. These findings are surprising, particularly given the higher prevalence of tobacco use disorder among Appalachian Kentuckians discharged during the study period, a known risk factor for IBD complications.
Although our findings show Appalachian Kentuckians with IBD have increased health care utilization, the reason for this increase is unclear. There does not appear to be a difference in IBD severity, as reflected by similar proportions of IBD-related complications between all cohorts evaluated. It is possible IBD-related complications are not accurately captured in the database due to inadequate coding, and more severe IBD in this population is still to blame. However, an increased proportion of surgical procedures among Crohn’s patients in Appalachian Kentucky would be expected in this scenario, yet this was not appreciated in the data set. We suspect increased health care utilization in the Appalachian region may reflect an increased incidence of IBD, rather than increased disease severity. While rural residence, tobacco, and opioid use are risk factors for increased health care utilization, the Kentucky Appalachian cohort had higher health care utilization compared with the national rural cohort. Furthermore, despite a higher proportion of opioid use disorder in Appalachian Kentucky, the rates were comparable to the national cohorts. This suggests there are additional risk factors impacting disease outcomes in this population, such as environmental exposures, dietary intake, dysbiosis, and increased genetic risk. Identification of additional risk factors is warranted, particularly as intervention targeting modification of these risk factors in vulnerable counties may be critical for disease prevention.
Finally, though Appalachian Kentuckians were more likely to require hospital admission, the number of hospital admissions trended downward during the period examined. The number of outpatient and ED visits increased. Unfortunately, we are unable to delineate outpatient visits by type of provider seen (gastroenterologist, PCP, surgeon, etc.), nor was prescription data available from either data source. The uptrend in ED visits in the Kentucky Appalachian cohort suggests IBD management was not optimized, perhaps due to lack of access to IBD specialty care. Further study utilizing alternate data sources that include outpatient prescription records will better clarify the disparity between these populations.
Limitations
This study is strengthened by the size of the Kentucky cohorts and by its comparison to robust national data; however, there are several limitations we must acknowledge. Due to the nature of the database, which comprises encounter-level data with no unique patient identifiers, it was impossible to follow patients over time and delineate individuals who had multiple readmissions. Therefore, we cannot exclude the possibility that a few high utilizers skewed the data sample, nor can we exclude the possibility that single encounters may have been miscoded. However, the use of gastrointestinal-related MS-DRG codes as part of the inclusion criteria for inpatient encounters lowers the probability of miscoding of those encounter types. We were also unable to reliably calculate a Charlson comorbidity index from these databases and therefore were unable to control for baseline comorbidities. We cannot exclude comorbid conditions as significant contributors to increased health care utilization in the Appalachian Kentucky cohort. The use of 1 year (2019) to compare state and national level data may also be a limitation, given that our findings may not reflect ongoing national trends. Finally, as mentioned previously, the accuracy of the data analyzed is subject to coding bias, including under coding and miscoding of diagnoses and procedures. It is possible certain IBD-related complications were incorrectly omitted from hospital coding and therefore not fully captured in this analysis.
Conclusion
Appalachian Kentuckians with IBD, particularly Crohn’s disease, are more likely to be high utilizers of health care than their non-Appalachian IBD counterparts. This is evidenced by higher rates of inpatient admissions, ER visits, and outpatient visits. They are significantly higher utilizers compared with the national IBD cohort, including the rural national cohort, which implies rural residence does not explain this disparity. This study confirms a disparity in health care utilization among Kentuckians with IBD and suggests disease incidence but not severity is greater in Appalachia. These findings may challenge the notion of a “rural protective” effect in rural areas. Future areas of study to further characterize and explain this finding will include analysis of patient-level data to address some of the limitations of the current study and provide estimates of prevalence. Additional avenues of investigation for this study may include identification of modifiable disease risk factors, including environmental exposure and dietary intake, particularly relating to differences in microbiome. Pharmacologic causes for disparate quality indicators of IBD care in the Appalachian region will also be investigated, including prescription of and patient adherence to biologic prescriptions, as well as prescriber opioid and corticosteroid stewardship.
Supplementary Material
Acknowledgments
The authors would like to acknowledge the Kentucky Cabinet for Health and Family Services and the Office of Health Data and Analytics for their role in collection of data for and maintenance of Kentucky HIDOSD. The authors also thank Dr. Stephen Hanauer for manuscript review and editorial contributions.
Contributor Information
Christian N Rhudy, University of Kentucky Healthcare, Specialty Pharmacy and Infusion Services, Lexington, Kentucky, USA.
Courtney L Perry, University of Kentucky College of Medicine, Department of Medicine, Division of Digestive Diseases and Nutrition, Lexington, Kentucky, USA; University of Kentucky Healthcare, Specialty Pharmacy and Infusion Services, Lexington, Kentucky, USA.
Gregory S Hawk, University of Kentucky, Dr. Bing Zhang Department of Statistics, Lexington, Kentucky, USA.
Deborah R Flomenhoft, University of Kentucky College of Medicine, Department of Medicine, Division of Digestive Diseases and Nutrition, Lexington, Kentucky, USA.
Jeffery C Talbert, University of Kentucky College of Medicine, Division of Biomedical Informatics, Lexington, Kentucky, USA.
Terrence A Barrett, University of Kentucky College of Medicine, Department of Medicine, Division of Digestive Diseases and Nutrition, Lexington, Kentucky, USA.
Author Contributions
Conceptualization: C.L.P., T.A.B., D.F. Methodology: C.N.R., J.T., C.L.P. Formal analysis: C.N.R., G.S.H., C.L.P. Data acquisition: C.N.R., J.T. Writing—original draft: C.L.P. and C.N.R. Writing—review and editing: C.L.P., C.N.R., T.A.B., D.R., G.S.H., JT. Approval of final manuscript: all authors.
Funding
The project described was supported by the NIH National Center for Advancing Translational Sciences through grant number UL1TR001998. C.P. is supported by the National Institutes of Health, Grant TL1TR001997 of CCTS funding. The authors’ work is also supported by the National Institute of Diabetes and Digestive and Kidney Diseases of the National Institutes of Health under Award Numbers 2RO1 DK095662-10A1 and R21DK118954. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. This work is also supported by the Veteran’s administration Merit award 1I01CX001353-01A1. The content is solely the responsibility of the authors and does not necessarily represent the official views of the VA.
Conflicts of Interest
T.A.B. has consulted and received honoraria for speaker’s bureau activities for Takeda and AbbVie pharmaceutical companies. No other authors have disclosures to report.
References
- 1. Collaborators GBDIBD. The global, regional, and national burden of inflammatory bowel disease in 195 countries and territories, 1990-2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet Gastroenterol Hepatol. 2020;5(1):17-30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Dahlhamer JM, Zammitti EP, Ward BW, Wheaton AG, Croft JB.. Prevalence of inflammatory bowel disease among adults aged >/=18 years-United States, 2015. MMWR Morb Mortal Wkly Rep. 2016;65(42):1166-1169. [DOI] [PubMed] [Google Scholar]
- 3. Ye Y, Manne S, Treem WR, Bennett D.. Prevalence of inflammatory bowel disease in pediatric and adult populations: recent estimates from large national databases in the United States, 2007-2016. Inflamm Bowel Dis. 2020;26(4):619-625. [DOI] [PubMed] [Google Scholar]
- 4. Bernstein CN, Walld R, Marrie RA.. Social determinants of outcomes in inflammatory bowel disease. Am J Gastroenterol. 2020;115(12):2036-2046. [DOI] [PubMed] [Google Scholar]
- 5. Ananthakrishnan A. Epidemiology and risk factors for IBD. Nature Reviews Gastroenterol Hepatol. 2015;12(4):205-217. [DOI] [PubMed] [Google Scholar]
- 6. Colombel JF, Mahadevan U.. Inflammatory bowel disease 2017: innovations and changing paradigms. Gastroenterology. 2017;152(2):309-312. [DOI] [PubMed] [Google Scholar]
- 7. Lega S, Dubinsky MC.. What are the targets of inflammatory bowel disease management. Inflamm Bowel Dis. 2018;24(8):1670-1675. [DOI] [PubMed] [Google Scholar]
- 8. Burt RK, Craig RM, Milanetti F, et al. Autologous nonmyeloablative hematopoietic stem cell transplantation in patients with severe anti-TNF refractory Crohn disease: long-term follow-up. Blood. 2010;116(26):6123-6132. [DOI] [PubMed] [Google Scholar]
- 9. Behringer B, Friedell GH.. Appalachia: where place matters in health. Prev Chronic Dis. 2006;3(4):A113. [PMC free article] [PubMed] [Google Scholar]
- 10. Huttlinger K, Schaller-Ayers J, Lawson T.. Health care in Appalachia: a population-based approach. Public Health Nurs. 2004;21(2):103-110. [DOI] [PubMed] [Google Scholar]
- 11. Hosseinpoor AR, Bergen N, Mendis S, et al. Socioeconomic inequality in the prevalence of noncommunicable diseases in low- and middle-income countries: results from the World Health Survey. BMC Public Health. 2012;12:474. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Agborsangaya CB, Lau D, Lahtinen M, Cooke T, Johnson JA.. Multimorbidity prevalence and patterns across socioeconomic determinants: a cross-sectional survey. BMC Public Health. 2012;12:201. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Pollard KJ, Linda. The Appalachian Region: A Data Overview from the 2015-2019 American Community Survey Chartbook. arc.gov: Appalachian Regional Comission; 2021. [Google Scholar]
- 14. Benchimol EI, Kaplan GG, Otley AR, et al. Rural and urban residence during early life is associated with risk of inflammatory bowel disease: a population-based inception and birth cohort study. Am J Gastroenterol. 2017;112(9):1412-1422. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Benchimol EI, Kuenzig ME, Bernstein CN, et al. ; Canadian Gastro-Intestinal Epidemiology Consortium. Rural and urban disparities in the care of Canadian patients with inflammatory bowel disease: a population-based study. Clin Epidemiol. 2018;10:1613-1626. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Machado SR, Jayawardana S, Mossialos E, Vaduganathan M.. Physician density by specialty type in urban and rural counties in the US, 2010 to 2017. JAMA Netw Open. 2021;4(1):e2033994. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Hoogland AI, Hoogland CE, Bardach SH, Tarasenko YN, Schoenberg NE.. Health behaviors in rural Appalachia. South Med J. 2019;112(8):444-449. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Brown JL, Wenrich TR.. Intra-family role expectations and reluctance to change identified as key barriers to expanding vegetable consumption patterns during interactive family-based program for Appalachian low-income food preparers. J Acad Nutr Diet. 2012;112(8):1188-1200. [DOI] [PubMed] [Google Scholar]
- 19. Ananthakrishnan AN, Khalili H, Konijeti GG, et al. A prospective study of long-term intake of dietary fiber and risk of Crohn’s disease and ulcerative colitis. Gastroenterology. 2013;145(5):970-977. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Ananthakrishnan AN, Khalili H, Konijeti GG, et al. Long-term intake of dietary fat and risk of ulcerative colitis and Crohn’s disease. Gut. 2014;63(5):776-784. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Voutilainen M, Hutri-Kahonen N, Tossavainen P, et al. Low childhood high density lipoprotein cholesterol levels and subsequent risk for chronic inflammatory bowel disease. Dig Liver Dis. 2018;50(4):348-352. [DOI] [PubMed] [Google Scholar]
- 22. Xu F, Carlson SA, Liu Y, Greenlund KJ.. Prevalence of inflammatory bowel disease among medicare fee-for-service beneficiaries-United States, 2001-2018. MMWR Morb Mortal Wkly Rep. 2021;70(19):698-701. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Xu F, Carlson SA, Liu Y, Greenlund KJ.. Urban-rural differences in health care utilization for inflammatory bowel disease in the USA, 2017. Dig Dis Sci. 2022;67(8):3601-3611. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Benchimol EK, Bernstein C, Nguyen G, et al. Rural and urban disparities in the care of Canadian patients with inflammatory bowel disease: a population-based study. Clin Epidemiol. 2018;10:1613-1626. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Kuenzig ME, Benchimol EI, Lee L, et al. The impact of inflammatory bowel disease in Canada 2018: direct costs and health services utilization. J Can Assoc Gastroenterol. 2019;2(Suppl 1):S17-S33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Kentucky Cabinet for Health and Family Services-Office of Health Policy Kentucky Inpatient Hospitalization Claims Files, 2016-2019. Frankfort, KY. 2020. [Google Scholar]
- 27. Kentucky Cabinet for Health and Family Services-Office of Health Policy Kentucky Outpatient Services Claims Files, 2016-2019. Frankfort, KY. 2020. [Google Scholar]
- 28. Agency for Healthcare Research and Quality Healthcare Cost and Utilization Project (HCUP). HCUP Databases. Rockville, MD. 2022. [PubMed] [Google Scholar]
- 29. Appalachian Regional Commission (ARC) Subregions in Appalachia. Washington, DC. 2021. [Google Scholar]
- 30. National Center for Health Statistics (NCHS) NCHS Urban-Rural Classification Scheme for Counties. Hyattsville, MD: Centers for Disease Control and Prevention;2017. [Google Scholar]
- 31. Healthcare Cost and Utilization Project (HCUP) HCUP CCS-Services and Procedures. Agency for Healthcare Research and Quality. www.hcup-us.ahrq.gov/toolssoftware/ccs_svcsproc/ccssvcproc.jsp. Published 2021. Accessed.
- 32. Coffey R, Barrett M, Houchens B, Brady J, Moy E, Ho K, Andrews R.. Methods Applying AHRQ Quality Indicators to Healthcare Cost and Utilization Project (HCUP) Data for the Sixth (2008) National Healthcare Quality Report. 2008. [Google Scholar]
- 33. Census Bureau County Population Totals: 2010-2020. 2021. [Google Scholar]
- 34. Klein RJ, Schoenborn CA.. Age adjustment using the 2000 projected U.S. population. Healthy People 2000 Stat Notes. 200120):1-9. [PubMed] [Google Scholar]
- 35. Houchens R, Ross D, Elixhauser A.. Final Report on Calculating National Inpatient Sample (NIS) Variances for Data Years 2012 and Later. 2015. [Google Scholar]
- 36. Houchens R. Inferences with HCUP State Databases Final Report. 2010. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.