Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Jun 1.
Published in final edited form as: Spat Spatiotemporal Epidemiol. 2021 Mar 6;37:100414. doi: 10.1016/j.sste.2021.100414

Trends in Spatial Access to Colonoscopy in South Carolina, 2000-2014

Jan M Eberth 1,2,3,*, Whitney E Zahnd 1,*, Michele J Josey 1,2,3, Mario Schootman 4, Peiyin Hung 5, Janice C Probst 1,5
PMCID: PMC8575121  NIHMSID: NIHMS1685780  PMID: 33980409

Abstract

Background:

Colonoscopy use has increased since Medicare began covering screening for average-risk persons. Our objective was to describe changes in spatial access to colonoscopy in South Carolina (SC) between 2000 and 2014.

Methods:

Using data from the SC Ambulatory Surgery Database, we created annual ZIP Code Tabulation Area (ZCTA) spatial accessibility scores. We assessed changes in accessibility, colonoscopy supply, and potential demand, overall and by metropolitan designation. Spatial clustering was also explored.

Results:

Spatial accessibility decreased across both small rural and metropolitan ZCTAs but was significantly higher in metropolitan areas during the first part of the study period . The proportion of persons with no access to colonoscopy within 30 minutes increased over time but was consistently higher in small rural areas. Clusters of low accessibility grew over time.

Conclusions:

The supply of colonoscopy facilities decreased relative to the potential demand, and clusters of low access increased, indicating a contraction of services.

Introduction

Colorectal cancer (CRC) is the only cancer type with a recommended screening test (i.e., endoscopy) that can both prevent and detect cancer early (Winawer et al., 1993). Unlike other screening tests that require limited clinical expertise to implement, such as the Fecal Immunochemical Test (FIT) or Fecal Occult Blood Test (FOBT), endoscopy (sigmoidoscopy and colonoscopy) necessitates specialized training and a steady volume to maintain competence (Bhangu et al., 2012; Ko et al., 2014; Pace et al., 2017). Colonoscopies are a more complex procedure than sigmoidoscopy and allow for visualization (and biopsies, as needed) of more the colorectal tract than sigmoidoscopy. The goal of the National Colorectal Cancer Roundtable is for 80% of all recommended individuals to be screened per the United States Preventive Services Task Force (USPSTF) guidelines, which recommend regular testing using either FIT, FOBT, and colonoscopy modalities. Patients who choose to have a screening colonoscopy and those who seek follow-up after an abnormal finding from a FIT or FOBT all ultimately end up having a colonoscopy (80% in every community - National Colorectal Cancer Roundtable, n.d.; Bibbins-Domingo et al., 2016). Some studies estimate that if FIT-based initiatives were instituted, there would be enough colonoscopy providers to screen 80% of the eligible population (Joseph et al., 2016). However, it is unknown how access to colonoscopy providers varies spatially or temporally.

Traditionally, gastroenterologists perform most colonoscopies (65%), followed by surgeons and primary care providers (Baxter et al., 2012). During 2000-2010, some rural areas experienced a decline in gastroenterologist supply, which intensified the shortage of colonoscopy providers in rural communities (Eberth et al., 2018). The loss of colonoscopy providers in rural areas is critical as rural populations consistently experience higher overall and late-stage incidence CRC incidence rates and higher CRC mortality compared to their urban counterparts (Henley et al., 2017; Zahnd et al., 2018, 2017). The decline in supply is also particularly concerning given increasing demand for colonoscopy since Medicare began covering the costs of colonoscopy for average-risk beneficiaries in 2001 and as more of the “Baby Boomer” population (born between 1945 and 1965) has become eligible for CRC screening (Song and Ferris, 2018). Colonoscopy has been the preferred screening modality for CRC in the past two decades, as a popular alternative to FOBT and sigmoidoscopy (Klabunde et al., 2011; Meissner et al., 2006). Between 2000 and 2018, the proportion of adults who had a colonoscopy within the past 10 years increased from 28% to 61%, while FOBTs saw declines during the same period (National Cancer Institute, n.d.). Understanding spatial access to colonoscopy providers across rural and urban communities is an essential first step to inform policy actions. Longitudinally quantifying accessibility to providers at a more granular level using rigorous spatial methods can facilitate the development of tailored interventions in each community and guide policies to address trends in access.

To better monitor trends in the colonoscopy workforce over space and time during a period of major reimbursement changes, we conducted an exploratory study on the spatial access to colonoscopy providers in South Carolina (SC) over time. Specifically, we investigated both the supply and potential demand for screening colonoscopy in ZIP Code Tabulation Areas (ZCTAs) from 2000 to 2014 using the two-step floating catchment area method (2SFCA) (Wei and Fanui, 2003).

Methods

We examined annual spatial access to colonoscopy services across ZCTAs in SC at roughly 5-year intervals between 2000 and 2014 using the 2SFCA method, which considers both the supply and potential (or actual) demand for services at a local level. Data on the supply of colonoscopy services was obtained from the SC Ambulatory Surgery Database and data on potential demand was obtained from the U.S. Census Bureau’s Decennial Census and American Community Survey.

Data Sources

We used data from the SC Ambulatory Surgery Database, which includes data on ambulatory services provided regardless of patient age or payor type, in emergency departments, hospitals, and ambulatory surgery centers (except Veteran’s Health Administration facilities). To identify colonoscopy providers for each year, we used the SC Ambulatory Surgery Database to identify healthcare providers who performed a colonoscopy on individuals aged 50–74 for each respective year. Unique colonoscopy facility locations were geocoded using the Geographic Information Systems (GIS) software ArcGIS Pro 10.6.

We estimated the annual number of adults of recommended CRC screening age (50-74 years of age) in each ZCTA. For the years 2000, 2005, and 2010, we used the 2000 and 2010 Decennial Census counts, respectively. Before the launch of the ACS, there were no annual ZCTA-level population estimates, so we applied 2000 population counts for the years 2000 and 2005. For 2014, we used the 2012-2016 ACS estimates, as the midpoint for this estimate was the year of interest. These data consist of 407 ZCTAs in years 2000 and 2005, based upon the 2000 Decennial Census, and 424 ZCTAs for the years 2010-2014, based upon the 2010 Decennial Census. ZCTAs are developed every ten years based upon census block and ZIP code overlaps, with the most frequently occurring ZIP code within a census tract being assigned to that ZCTA. Thus, there are different number of ZCTAs and different boundaries during the 2000-2005 and the 2010-2014 periods of our study, respectively (U.S. Census Bureau, n.d.)

Two-Step Floating Catchment Area Method

We applied the 2SFCA method to create an annual spatial accessibility score for each ZCTA. The 2SFCA method is employed using Network Analyst in ArcGIS Pro. The details of this particular method are detailed elsewhere, but we provide an overview here. For the first step of the 2SFCA, we created a network-based 30-minute travel time catchment area around each facility. A 30-minute catchment area a commonly used metric to classify areas of low access to health care services that has also been frequently used when applying the 2SFCA method to determine access to cancer screening services or primary care within a single state or local area (Health Professional Shortage Areas (HPSAs) ∣ Bureau of Health Workforce, n.d.; Lian et al., 2012; Wei and Fanui, 2003).We then determined a provider-to-population ratio based upon the number of colonoscopy providers who practice at each facility (numerator) and the estimated population of people aged 50-74 (denominator) who lived in ZCTAs whose centroids were contained within the 30-minute catchment area. For example, if 5 providers performed colonoscopies at a particular facility and 100 persons of recommended screening age lived in ZCTAs whose centroids fell within the 30-minute catchment area of the facility, the provider-to-population ratio for that facility would be 5:100. Since some providers may perform colonoscopies at different facilities, we considered the proportion of colonoscopies performed at each location and weighted providers at a given facility accordingly. If a provider performed 100% of colonoscopies at a single location, they would be counted as “1” provider at that location. If, for example, they provided 75% of their colonoscopies at one location and 25% of their colonoscopies at another location, then they would be considered “0.75” provider and “0.25” provider at each location respectively. For step two of the 2SFCA, we constructed a network-based 30-minute catchment areas around each ZCTA centroid and summed the provider-to-population ratios of the locations that were contained within the 30-minute catchment area. This value was the spatial accessibility score for each ZCTA. Values can range from 0, indicating no spatial access to providers at a facility providing colonoscopies, to a hypothetical but unrealistic, value of 1, which indicates a ratio of 1 provider per person of recommended screening age.

Rural-Urban Measures

We used the ZCTA-approximated 2003 and 2013 US Department of Agriculture Rural-Urban Commuting Area (RUCA) primary codes to categorize ZCTAs as small rural, micropolitan, or metropolitan. RUCA codes were developed to characterize the population in census tracts based upon their metropolitan status and commuting patterns. The 2003 RUCA codes are based on the 2000 Census, and we applied these to the 2000 and 2005 spatial accessibility scores. The 2013 RUCA codes were developed from the 2010 Census, and we applied these to the 2010 and 2014 spatial accessibility scores. For both versions, RUCA codes were grouped with 1-3 as metropolitan, 4-6 as micropolitan, and 7-10 as small rural. Of note, the updates to the RUCA codes between the 2000 and 2010 Census led to a shift in the distribution of ZCTAs by metropolitan status (e.g., 36% of SC ZCTAs designated as small rural/micropolitan in 2003 were designated as metropolitan in 2013).

Statistical and Spatial Analysis

We calculated the overall percent change in the number of colonoscopy providers, facilities providing colonoscopy, potential demand for colonoscopy, provider-population ratio, and spatial accessibility scores from 2000 to 2014 for all ZCTAs stratified by metropolitan status. We performed Kruskal-Wallis test to compare median scores for metropolitan, micropolitan, and small rural areas by year.

We also examined spatial clustering of accessibility scores at roughly 5-year increments (2000, 2005, 2010, and 2014) by calculating Global Moran’s I statistics, Local Moran’s I test also known as the Local Indicators of Spatial Autocorrelation (LISA) analysis, and LISA transitions using ArcGIS Pro. Global Moran’s I is a statistic that evaluates spatial autocorrelation, which assesses how similar values (e.g., spatial accessibility scores) are to other nearby values (Spatial Autocorrelation (Global Moran’s I) (Spatial Statistics)—ArcGIS Pro ∣ Documentation, n.d.). Values of this statistic range from −1 (perfect dispersion) to +1 (perfect clustering) with a value of 0 representing a random spatial pattern. The LISA analysis can identify where there are areas where high and low access are clustered (hot and cold spots, respectively), as well as identity outliers (Anselin, 1995). Potential values included: high-high clusters, low-low clusters, high-low outliers, low-high outlier, and non-significant. High-high clusters indicate areas where there are areas of high values clustered together; low-low clusters are areas where low values are clustered. High-low Outliers are areas with high values surrounded by low values, while low-high outliers indicate areas of low values surrounded by areas of high values. For these analyses, we conceptualized the spatial relationship between ZCTAs using inverse distance (e.g., nearby features have a larger influence on the analysis than those far away), and the spatial weights were row standardized. We also performed a LISA cluster transition analysis to assess how ZCTAs changed between 2000 and 2005 and between 2010 and 2014, respectively. We used these comparison points due to the change in ZCTA boundaries/frequencies following the 2010 Decennial Census. The LISA transitions analysis shows how ZCTAs change in their clustering patterns across time (e.g., a change from a high-high cluster to a non-significant value) or if the clustering pattern was constant across time (Martin et al., 2017). Statistical significance was based on an alpha level of 0.05.

Sensitivity Analysis

To assess the dynamics of changes in provider counts, facility locations, and populations, we also performed a sensitivity analysis to assess the counterfactual scenario of how spatial accessibility scores may be different in 2014 had the number and locations of colonoscopy providers been the same as in 2000 while considering the changing population. Under this counterfactual scenario, we calculated spatial accessibility scores across ZCTAs, performed a LISA analysis, and compared the median spatial accessibility scores across small rural, micropolitan, and metropolitan designations.

Results

The number of colonoscopy screening locations in SC increased from 72 locations in 2000 to 91 locations in 2014 (26% increase) (Figure 1A). However, the number of locations decreased by 31% in small rural ZCTAs (13 in 2000 to 9 in 2014) and by 22% in micropolitan ZCTAs (18 in 2000 to 14 in 2014) while the number increased 66% in metropolitan ZCTAs (41 in 2000 to 68 in 2014). The number of unique colonoscopy providers increased from 883 providers in 2000 to 1,069 providers in 2014 (21% increase; Figure 1B), which translates to an increase of 77% in small rural, 9% in micropolitan, and 19% increase in metropolitan ZCTAs. The number of individuals of recommended screening age increased by 59% from 905,422 in 2000 to 1,436,751 in 2014 (Figure 1C), which included a 24% decrease in small rural residents of recommended screening age and an 89% increase in eligible metropolitan residents, while the micropolitan population grew 7%. The provider-population ratio decreased 24% from 2000 to 2014 (97.6 to 74.4 per 100,000), which included a 133% increase in small rural areas, a 2% increase in micropolitan areas, and a 37% reduction in metropolitan areas (Figure 1D).

Figure 1.

Figure 1.

Open in a new tab

Trends in Supply and Demand for Colonoscopy, 2000-2014 (A: Providers, B: Locations, C: Population of Recommended Screening Age; D: Provider-Population Ratio)

Spatial accessibility scores decreased by 13% across ZCTAs from 2000 to 2014. Median spatial accessibility decreased in all groups: 21% in metropolitan, 37% in micropolitan, and 18% in small rural. The proportion of the screening-eligible population living ZCTAs with no access to colonoscopy services within 30 minutes travel time remained relatively stable between 2000 and 2014 for metropolitan areas (Figure 2B). Among small rural ZCTAs, the proportion of the screening-eligible population living ZCTAs with no access to colonoscopy within 30 minutes increased from 7% to 13.9% during the study period. The proportion of micropolitan ZCTAs with no access increased from 1.5% to 5.2%.

Figure 2.

Figure 2.

Open in a new tab

Trends in Spatial Accessibility Score by Metropolitan, Micropolitan, and Small Rural Designations (A: Median Scores, B: Proportion of Screening Eligible Populations in ZIP Code Tabulation Areas with No Access to Colonoscopy with 30 Minutes’ Travel Time)

Moderate spatial clustering of colonoscopy accessibility scores in SC ZCTAs was evident across all years studied (Global Moran’s I range = 0.557-0.628, p<0.0001). Clusters of high access (in blue in Figure 3) to colonoscopy persisted across all years in the metropolitan areas of Greenville and Charleston, albeit the cluster shrunk in Charleston, as well as in the more rural northeastern section of the state. A cluster of high access emerged in the metropolitan area of Columbia (center of the state) in 2014. Clusters of low spatial access (in red) grew larger over time and were largely situated in more rural areas of SC (Figure 3). Rural communities along the I-95 Coordinator in the Lowcountry had increasingly low spatial access to colonoscopy providers (i.e., represented in the dark red areas in Figure 3) over time. Our LISA transition analysis showed that areas of consistent high-high clusters were in the Charleston area on the coast and in the northeastern part of the state and in the Greenville area in the Upstate in both 2000 and 2005 (Figure 2A). Persistent low-low clusters and emerging low-low clusters were found in the PeeDee region in the eastern part of the state and in the west-central part of the state. Figure 2B shows the transition in clusters between 2010 and 2014. Like the earlier LISA transitions, areas of high-high clusters persisted in the Greenville area in the Upstate of South Carolina, the northeastern part of the state, and in the Charleston area. High-high clusters emerged in the Columbia area in the middle of the state. Areas of persistent low-low clusters were in the western parts of the state as well as further northeast in the PeeDee region of the state.

Figure 3:

Figure 3:

Open in a new tab

Local Indicators of Spatial Autocorrelation (LISA) of Spatial Accessibility Scores in South Carolina ZIP Code Tabulation Areas (A: 2000, B: 2005, C: 2010, D:2014)

Our sensitivity analysis showed that the spatial clustering of spatial accessibility scores under the counterfactual scenario for the 2014 population was largely similar to the main scenario with somewhat more diffuse areas of both cold and hotspots (Supplementary Figure 1). There were some differences in median spatial accessibility scores, with the scores being slightly lower in the counterfactual scenario overall and among metropolitan ZCTAs (Supplementary Table 1). While the differences between the 2014 median scores and the counterfactual scores were minor, the counterfactual scenario yields twice as many ZCTAs with no access within 30 minutes (7.4% of ZCTAs in the counterfactual scenario vs. 3.7% of ZCTAs in the main analysis).

Discussion

We examined trends in spatial access to colonoscopy between 2000 and 2014, which revealed a decreasing trend in spatial access to care throughout the state, despite an overall increase in the number of colonoscopy locations and providers over the study period. The increase in the number of colonoscopy locations and providers did not result in increased accessibility to CRC screening at the local level because of the disproportionate growth of providers in metropolitan areas and the overall speed of population growth during this time. Areas of higher access existed in the most densely populated metropolitan areas of the state.

Despite an increasing colonoscopy provider supply during 2000-2014, accessibility to colonoscopy decreased over time considering the potential demand, implying an unbalanced geographic redistribution of colonoscopy providers in SC communities. Our results reveal that some areas, particularly micropolitan and small rural areas, experienced persistently lower access throughout the 15-year study period. In contrast, several clusters with high access to colonoscopy services were in metropolitan areas. Such maldistribution of facilities in which at-risk residents could lack access to colonoscopy may hinder ongoing statewide and national efforts to reach the National Colorectal Cancer Roundtable’s goal of “80% in Every Community,” particularly if colonoscopy remains our primary CRC screening modality (80% in every community - National Colorectal Cancer Roundtable, n.d.).

Our findings on decreasing spatial accessibility, while clusters of high access grew in the some of the most metropolitan areas of the state have a few implications. First, while the number of providers and locations grew over time, they increasingly practiced in the most metropolitan areas, contributing to clusters of high spatial access in those areas. Second, the demand for colonoscopy services increased at a higher rate—though not in small rural areas—than the increase in the number of providers or locations. Both corroborate previous findings (Eberth et al., 2018). Our use of Decennial Census and ACS population estimates allowed us to consider the population dynamics of the growth of “Baby Boomers” (born between 1945 and 1965) who aged into the CRC screening cohort during the study period. This large aging population has created health care capacity challenges due to the divergence in supply and capacity of health care services and the demand for healthcare services. (Retooling for an aging America: Building the health care workforce, 2008; Song and Ferris, 2018). Our findings showed some unexpected dynamics in supply and demand. While aging “Baby Boomers” increased the potential demand for colonoscopy in metropolitan areas, the absolute number of persons of recommended screening age decreased in both micropolitan and small rural areas over the study period. Meanwhile, in small rural areas, the number of facilities providing colonoscopies decreased while the number of unique providers increased substantially over the 2000-2014 study period (increasing the provider-population ratio). This trend may be due to, in part, to increasing consolidation in gastroenterology care due to increased regulatory burden, low reimbursements, and low profit margins disproportionately affecting small community practices (Allen, 2012). Some of this mixed dynamic may be due, in part, to changes in commuting patterns and demographic shifts between the 2000 and 2010 Decennial Census that may contribute to changes in rural-urban designations in RUCAs. However, our findings are consistent with county-based demographic analyses in which the rural-urban designation was held constant that also identified shrinking rural populations (Thiede et al., 2017). On the other hand, small rural areas have disproportionately more older adults than metropolitan areas, a larger percentage is older than age 75, the upper age range of individuals recommended for CRC screening, indicating that some of the reduced “demand” for screening in rural may be due in part more rural persons exceeding the recommended screening age. Consistent with previous studies, the number of primary care physicians providing colonoscopy in SC increased over time, contributing to the overall growth in colonoscopy providers in small rural areas (Rosenblatt and Hart, 2000). The dynamic between provider type/volume, population demographics, and geographic location remains an important area of future research.

We found that spatial access to colonoscopy providers was persistently lower in rural populations and that the number of facilities providing colonoscopy in rural areas decreased. The persistent and widening lack of spatial access to colonoscopy facilities in small rural/micropolitan areas may be due to, in part, to the increase in ambulatory care centers performing colonoscopy in SC and other states (Eberth et al., 2018; Hollingsworth et al., 2011). Of note, rural ambulatory surgery centers are less prevalent in states with certificate of need (CON) policies compared to those without such policies (Gregg et al., 1997). As SC has the 9th most restrictive CON program in the U.S., these limitations may reduce access to critical health services, including preventive care like colonoscopy (Koopman et al., 2015).

With the decrease in access to colonoscopy providers, expansion of other proven CRC screening modalities such as FIT or office-based sigmoidoscopy represent an opportunity to ensure adequate access to CRC screening in areas where there is less access to colonoscopy. Some have suggested that a national program with FIT as the primary screening test with follow-up colonoscopy (vs. first-line screening) would provide enough capacity to screen 80% of the eligible U.S. population. For ZCTAs where spatial accessibility to colonoscopy providers has been consistently lower, state governments and community stakeholders may develop community health programs, mobile health screening, and promotion of other CRC screening modalities. Specifically, FIT-based interventions have been shown to be effective in rural communities where colonoscopy is lacking (Crosby et al., 2017).

Our study was not without limitations. First, as we only had access to and included colonoscopy providers in SC, spatial accessibility scores for ZCTAs closer to the Georgia and North Carolina borders may be underestimated or overestimated due to edge effects, as residents of neighboring states may have a colonoscopy in SC and its residents may have a colonoscopy in a neighboring state. We used ZCTAs as our geographic unit which is susceptible to spatial mismatch and discontinguity (Grubesic and Matisziw, 2006). ZCTAs are also developed by the U.S. Census Bureau at each Decennial Census. Thus, there are incongruencies in the number and shapes of ZCTAs in temporal studies, such as ours, that cross Decennial Census years which may lead to modifiable areal unit problem/zonal effect concerns that impact study findings. We used the American Community Survey to characterize demand for colonoscopy services by estimating the screening eligible population; however, 2014 ACS data are intended to describe rather than estimate population groups. Unlike the Decennial Census, the ACS uses a sampling design, and therefore is susceptible to both sampling and non-sampling errors especially within small areas (Spielman et al., 2014). Further, since annual population estimates were not available prior to the 2007 initiation of the ACS, we had to use somewhat dated population estimates for 2005. Additionally, we used Decennial Census and ACS population estimates for persons 50-74 years of age, as those were the age groupings available in the ACS and census data that were most congruent with USPSTF recommendations. However, persons at high risk may have a screening colonoscopy at ages younger than 50 and/or persons may continue to have a colonoscopy at ages older than USPSTF recommendations. Our results show, even without these distinct cases included, the supply has not kept up with the needed demand over time.

Despite these limitations, our study had several strengths. First, we used an administrative, all-payer database to determine the location and the number of providers who performed colonoscopies among those of recommended screening age. Second, we used the 2SFCA method, rather than densities or distance, which enable us to better examine the supply of and demand for colonoscopy services while also considering travel time. Previous studies have suggested that spatial clustering of access to cancer screening services and cancer outcomes be used to determine where interventions may be most needed (Lofters et al., 2013; Siegel et al., 2015). Workforce trends analyses should be explicitly structured to identify regional and subgroup variation that will inform where to target limited resources.

Conclusion

Despite the increased number of physicians and facilities providing colonoscopies during our study period, spatial access to colonoscopy has decreased over time with persistently low access in rural areas. Although national simulation studies suggest that there is a sufficient workforce to ensure goals for CRC screening are met, future research should examine how the available workforce varies across regions and across the rural-urban continuum. Additionally, research should not only consider changes in supply, but also changes in demand—especially as the even the youngest of the populous “Baby Boomer” generation remains of recommended screening age until 2040.

Supplementary Material

1

Figure 4:

Figure 4:

Open in a new tab

Local Indicators of Spatial Autocorrelation (LISA) Transitions of Spatial Accessibility Scores in South Carolina ZIP Code Tabulation Areas: A: 2000 to 2005; B: 2010 to 2014

Table 1.

Median Spatial Accessibility to Colonoscopy Providers in South Carolina ZCTAs, Overall and by Rural-Urban Designation, 2000, 2005, 2010 and 2014

Overall Small
Rural		 Micropolitan Metropolitan % Difference
(Urban vs. Small
Rural)		 % Difference (Urban
vs. Micropolitan)		 P value*
2000 0.00063 0.000474 0.000602 0.000856 81% 42% 0.002
2005 0.000665 0.000386 0.000665 0.000761 97% 14% <0.001
2010 0.00061 0.00049 0.000469 0.000639 30% 36% 0.58
2014 0.000553 0.000385
0.000391		 0.000384 0.000673 72% 75% 0.34
Open in a new tab
*

Kruskal-Wallis Test comparing urban, micropolitan, and rural ZCTAs.

Highlights.

  • Access to colonoscopy is important to reduce the colorectal cancer burden.

  • Spatial access to colonoscopy decreased in rural and urban South Carolina.

  • The increase in the screening-eligible population contributed to decreased access.

  • Urban contraction of colonoscopy providers also contributed to decreased access.

  • Spatial clusters of low access grew over time.

Acknowledgements:

This publication was also made possible in part by grant numbers T32-GM081740 from NIH-National Institute of General Medical Sciences (Josey), MRSG-15-148-01-CPHPS from the American Cancer Society (Eberth, Probst, Schootman), and 5U1CRH0311-12-00 from the Federal Office of Rural Health Policy (Eberth, Probst). The contents are those of the author(s) and do not necessarily represent the official views of, nor an endorsement, by NIH, HRSA, HHS, or the U.S. Government.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Conflicts of Interest: The authors declare no conflicts of interest.

References

  1. 80% in every community - National Colorectal Cancer Roundtable. n.d. https://nccrt.org/80-in-every-community/ (accessed March 28, 2020).
  2. American Lung Association. Cutting Tobacco’s Rural Roots: Tobacco Use in Rural Communities. http://www.lung.org/assets/documents/research/cutting-tobaccos-rural-roots.pdf. (accessed January 28, 2021).
  3. Allen JI. The Road Ahead. Clinical Gastroenterology and Hepatology 2012;10:692–6. 10.1016/j.cgh.2012.05.002. [DOI] [PubMed] [Google Scholar]
  4. Anselin L Local Indicators of Spatial Association—LISA. Geographical Analysis 1995;27:93–115. 10.1111/j.1538-4632.1995.tb00338.x. [DOI] [Google Scholar]
  5. Baxter NN, Warren JL, Barrett MJ, Stukel TA, Doria-Rose VP. Association between colonoscopy and colorectal cancer mortality in a US cohort according to site of cancer and colonoscopist specialty. J Clin Oncol 2012;30:2664–9. 10.1200/jco.2011.40.4772. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bhangu A, Bowley DM, Horner R, Baranowski E, Raman S, Karandikar S. Volume and accreditation, but not specialty, affect quality standards in colonoscopy. British Journal of Surgery 2012;99:1436–44. 10.1002/bjs.8866. [DOI] [PubMed] [Google Scholar]
  7. Bibbins-Domingo K, Grossman DC, Curry SJ, Davidson KW, Epling JW, García FAR, et al. Screening for colorectal cancer: US preventive services task force recommendation statement. JAMA - Journal of the American Medical Association 2016;315:2564–75. 10.1001/jama.2016.5989. [DOI] [PubMed] [Google Scholar]
  8. Crosby RA, Stradtman L, Collins T, Vanderpool R. Community-based colorectal cancer screening in a rural population: who returns Fecal Immunochemical Test (FIT) kits? J Rural Health 2017;33:371–4. 10.1111/jrh.12210. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Eberth JM, Josey MJ, Mobley LR, Nicholas DO, Jeffe DB, Odahowski C, et al. Who Performs Colonoscopy? Workforce Trends Over Space and Time. J Rural Health 2018;34:138–47. 10.1111/jrh.12286. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Gregg W, Douglas Wholey M, Moscovice I. The Impact of Freestanding Ambulatory Surgery Centers on Rural Community Hospital Performance, 1997–2006. 1997. [Google Scholar]
  11. Grubesic TH, Matisziw TC. On the use of ZIP codes and ZIP code tabulation areas (ZCTAs) for the spatial analysis of epidemiological data. International Journal of Health Geographics 2006;5. 10.1186/1476-072X-5-58. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Health Professional Shortage Areas (HPSAs) ∣ Bureau of Health Workforce. n.d. https://bhw.hrsa.gov/shortage-designation/hpsas (accessed January 19, 2020).
  13. Henley SJ, Anderson RN, Thomas CC, Massetti GM, Peaker B, Richardson LC. Invasive Cancer Incidence, 2004-2013, and Deaths, 2006-2015, in Nonmetropolitan and Metropolitan Counties - United States. MMWR Surveill Summ 2017;66:1–13. 10.15585/mmwr.ss6614a1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Hollingsworth JM, Krein SL, Ye Z, Kim HM, Hollenbeck BK. Opening of ambulatory surgery centers and procedure use in elderly patients: Data from Florida. Archives of Surgery 2011;146:187–93. 10.1001/archsurg.2010.335. [DOI] [PubMed] [Google Scholar]
  15. Joseph DA, Meester RGS, Zauber AG, Manninen DL, Winges L, Dong FB, et al. Colorectal cancer screening: Estimated future colonoscopy need and current volume and capacity. Cancer 2016;122:2479–86. 10.1002/cncr.30070. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Klabunde CN, Cronin KA, Breen N, Waldron WR, Ambs AH, Nadel MR. Trends in colorectal cancer test use among vulnerable populations in the United States. Cancer Epidemiology Biomarkers and Prevention 2011;20:1611–21. 10.1158/1055-9965.EPI-11-0220. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Ko CW, Dominitz JA, Neradilek M, Polissar N, Green P, Kreuter W, et al. Determination of Colonoscopy Indication From Administrative Claims Data. Medical Care 2014;52:e21–9. 10.1097/MLR.0b013e31824ebdf5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Koopman C, Stratmann T, Elbarasse M. MERCATUS ON POLICY Certificate-of-Need Laws: Implications for South Carolina. 2015. [Google Scholar]
  19. Lian M, Struthers J, Schootman M. Comparing GIS-based measures in access to mammography and their validity in predicting neighborhood risk of late-stage breast cancer. PLoS One 2012;7:e43000. 10.1371/journal.pone.0043000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Lofters AK, Gozdyra P, Lobb R. Using geographic methods to inform cancer screening interventions for South Asians in Ontario, Canada. BMC Public Health 2013;13:395. 10.1186/1471-2458-13-395. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Martin MJR, Matthews SA, Lee BA. The Spatial Diffusion of Racial and Ethnic Diversity Across U.S. Counties. Spatial Demography 2017;5:145–69. 10.1007/s40980-016-0030-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Meissner HI, Breen N, Klabunde CN, Vernon SW. Patterns of colorectal cancer screening uptake among men and women in the United States. Cancer Epidemiology Biomarkers and Prevention 2006;15:389–94. 10.1158/1055-9965.EPI-05-0678. [DOI] [PubMed] [Google Scholar]
  23. National Cancer Institute. Colorectal Cancer Screening ∣ Cancer Trends Progress Report https://progressreport.cancer.gov/detection/colorectal_cancer (accessed April 19, 2020). [Google Scholar]
  24. Pace D, Borgaonkar M, Evans B, Marcoux C, Lougheed M, Falk V, et al. Annual colonoscopy volume and maintenance of competency for surgeons. Surgical Endoscopy 2017;31:2630–5. 10.1007/s00464-016-5275-1. [DOI] [PubMed] [Google Scholar]
  25. Retooling for an aging America: Building the health care workforce. National Academies Press; 2008. 10.17226/12089. [DOI] [PubMed] [Google Scholar]
  26. Rosenblatt RA, Hart LG. Physicians and rural America. Western Journal of Medicine 2000;173:348–51. 10.1136/ewjm.173.5.348. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Siegel RL, Sahar L, Robbins A, Jemal A. Where can colorectal cancer screening interventions have the most impact? Cancer Epidemiol Biomarkers Prev 2015;24:1151–6. 10.1158/1055-9965.EPI-15-0082. [DOI] [PubMed] [Google Scholar]
  28. Song Z, Ferris TG. Baby Boomers and Beds: a Demographic Challenge for the Ages. Journal of General Internal Medicine 2018;33:367–9. 10.1007/s11606-017-4257-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Spatial Autocorrelation (Global Moran’s I) (Spatial Statistics)—ArcGIS Pro ∣ Documentation. https://pro.arcgis.com/en/pro-app/tool-reference/spatial-statistics/spatial-autocorrelation.htm (accessed November 29, 2020). [Google Scholar]
  30. Spielman SE, Folch D, Nagle N. Patterns and causes of uncertainty in the American Community Survey. Applied Geography 2014;46:147–57. 10.1016/j.apgeog.2013.11.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Thiede BC, Brown DL, Sanders SR, Glasgow N, Kulcsar LJ. A Demographic Deficit? Local Population Aging and Access to Services in Rural America, 1990–2010. Rural Sociology 2017;82:44–74. 10.1111/ruso.12117. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. USDA ERS - Rural - Urban Commuting Area Codes. https://www.ers.usda.gov/data-products/rural-urban-commuting-area-codes.aspx (accessed November 29, 2020).
  33. Wei L, Fanui LWW. Measures of Spatial Accessibility to Health Care in a GIS Environment: Synthesis and a Case Study in the Chicago Region 2003;30. 10.1068/b29120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Winawer SJ, Zauber AG, Ho MN, O’Brien MJ, Gottlieb LS, Sternberg SS, et al. Prevention of Colorectal Cancer by Colonoscopic Polypectomy. New England Journal of Medicine 1993;329:1977–81. 10.1056/NEJM199312303292701. [DOI] [PubMed] [Google Scholar]
  35. Zahnd WE, Fogleman AJ, Jenkins WD. Rural–Urban Disparities in Stage of Diagnosis Among Cancers With Preventive Opportunities. American Journal of Preventive Medicine 2018;54:688–98. 10.1016/j.amepre.2018.01.021. [DOI] [PubMed] [Google Scholar]
  36. Zahnd WE, James AS, Jenkins WD, Izadi SR, Fogleman AJ, Steward DE, et al. Rural-Urban Differences in Cancer Incidence and Trends in the United States. Cancer Epidemiol Biomarkers Prev 2017;27:1265–74. 10.1158/1055-9965.epi-17-0430. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. ZIP Code Tabulation Areas (ZCTAs). https://www.census.gov/programs-surveys/geography/guidance/geo-areas/zctas.html (accessed January 23, 2021). [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1685780-supplement-1.docx (161.9KB, docx)

RESOURCES