Utilization Pattern of Computed Tomographic Colonography in the United States: Analysis of the U.S. National Health Interview Survey

Colorectal cancer is the second leading cause of cancer deaths in the United States (1, 2). In 2017 alone, 135,430 individuals were diagnosed with colorectal cancer, and 50,260 individuals died from colorectal cancer (1). Colorectal cancer is one of the few cancers, which is preventable and also curable if diagnosed in early stages (3). However, more than one-third of individuals who are eligible for screening are not up-to-date with colorectal cancer screening (4, 5). The U.S. Preventive Services Task Force (USPSTF) recommends screening for colorectal cancer in average-risk individuals ages 50–75 years (3), although recent guidelines from the American Cancer Society (ACS) in 2018 suggest initiating screening earlier at age 45 (6). Colonoscopy is the most commonly used screening modality for colorectal cancer—more than 95% of those who underwent colorectal cancer screening elected to use colonoscopy (4, 7). Despite its demonstrated clinical benefits, individuals often delay or forego colonoscopy due to its relatively expensive cost, invasiveness, and other barriers (e.g., bowel cleanse); these characteristics are considered to impede optimal screening compliance rates for colorectal cancer (8, 9).

One way to increase screening is to offer alternative screening strategies that are less invasive and more cost-effective. CT colonography (also known as “virtual colonoscopy”) is one of the alternative tests developed for colorectal cancer screening (10, 11). The technique combines helical CT scanning and three-dimensional image rendering of the cleansed, enlarged colorectum replicating the view of the conventional colonoscopy (11). CT colonography uses this advanced visualization, allowing a minimally invasive structural evaluation of the entire colorectum and also provides details of other abdominal structures. Studies have demonstrated comparable clinical and economic benefits of CT colonography to those of colonoscopy (12–14). For example, in a landmark trial (15), CT colonography showed 89%–94% sensitivity and 80%–96% specificity in the detection of adenomatous polyps (≥6 mm), which was comparable with those obtained with optical colonoscopy. Although its low specificity for small polyps (<6 mm) has been a concern (16), CT colonography has shown promising results when compared with other colorectal cancer screening modalities. Graser and colleagues (17) demonstrated that the sensitivity of CT colonography (96.7%) for detecting advanced adenomas was comparable with that of colonoscopy (close to 100%) and also superior to sigmoidoscopy and fecal tests (range of 20%–83.3%). Notably, it appears that CT colonography has a better patient acceptance rate and a lower risk of procedure-related complications compared to other screening tests (18, 19).

Although accruing evidence suggests that CT colonography has the potentials to address the limitations of existing screening tools and improve low compliance with colorectal cancer screening, little is known about utilization patterns and individual factors for choosing CT colonography for colorectal cancer screening. Previous studies estimated its limited use of <2% for colorectal cancer screening among individuals ages 55–75 (7, 20). One prior study (20) found that insurance coverage for CT colonography was variable (e.g., Medicare did not cover CT colonography); nevertheless, having coverage for CT colonography was the primary predictor of CT colonography. In this regard, existing evidence is limited because most of them were conducted before the Affordable Care Act (ACA), which has greatly expanded access to colorectal cancer screening (21). Considering increased coverage for cancer screening under the ACA (21–23), large population-based studies using recent data are needed to examine access to CT colonography and better inform national investment to improve colorectal cancer screening and surveillance.

As such, this study analyzed a nationally representative survey data to examine patterns of recent CT colonography utilization and to identify the individual characteristics associated with using CT colonography for colorectal cancer screening.

This study was a retrospective cross-sectional analysis of the National Health Interview Survey (NHIS) 2015 and 2018. The NHIS is an annual household survey of noninstitutionalized individuals that uses a multistage probability design to capture a broad geographic representation of the United States (average response rate of >80%; ref. 24). Because first administered in 1957, the NHIS has been the main source of data to monitor population health patterns and trends in the United States. Of particular interest in this study, we used data from the NHIS' Cancer Control Supplement (CCS), which is sponsored by the National Cancer Institute. The NHIS-CCS collects additional cancer-related information, including knowledge, attitudes, screening, and risk assessment (25). We relied on the data from the NHIS-CCS 2018 and 2015 because they are the most recent data that contain a subset of cancer screening questions (25). This study was conducted in accordance with the ethical guidelines of the Belmont Report, and deemed exempt from review by the University of Florida Institutional Review Board because of the nature of publicly available data. Thus, no informed consent was required. We followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline (26).

The study population included U.S. adults ages 45 years or older (top coded at age 85+) without a diagnosis of colorectal cancer. The current USPSTF guidelines recommend colorectal cancer screening starting at age 50 and continuing until 75 (3). We opted to include those ages 45–49 and 76–85 to further examine utilization patterns of CT colonography among those not age eligible (e.g., often excluded from analysis), considering the recent update to the ACS guidelines to lower the age of screening initiation (6) and promoted shared screening decision making among those ages 75+ (27). We only excluded those with missing information on health insurance, education, and employment status (<0.8% of the total sample). Supplementary Table S1 presents the details of the study cohort selection process.

Receipt of CT colonography was our primary variable for this study. NHIS respondents self-reported their history of screening tests for colorectal cancer, including colonoscopy, sigmoidoscopy, fecal occult blood tests (FOBT), and CT colonography. Specifically, respondents were queried with the following question, “When did you have your MOST RECENT CT colonography or virtual colonoscopy?” The use of other screening modalities was also measured with parallel questions, “When did you have your MOST RECENT [name of screening tests: colonoscopy, sigmoidoscopy or home blood stool test]?” To assess the overall utilization of CT colonography, we included those who reported ever had CT colonography without a specific time interval. Then, we restricted our analysis to those identified as being up-to-date on colorectal cancer screening with any modality. On the basis of the current guidelines, colorectal cancer screening up-to-date status was defined as having any of colorectal cancer screening tests within a recommended time period: colonoscopy within 10 years, sigmoidoscopy within 5 years, FOBT within the past year, and CT colonography within 5 years (3, 6).

We included study participants' sociodemographic characteristics to examine their associations with CT colonography use. Those included were: age (45–49, 50–64, 65–75, and 76+ years), sex, race/ethnicity (non-Hispanic White, non-Hispanic Black, Hispanic, and Other races), marital status, employment, education attainment [less than high school, high school graduate or general equivalency diploma (GED), some college, and Bachelor's degree or higher], family income [defined by federal poverty level (FPL); low income, FPL < 200%, middle income, FPL 200%–400%, and high income, FPL > 400%], and type of health insurance (any private, public, and uninsured).

In all analyses, weighted frequencies and prevalence with 95% confidence interval (CI) were calculated to report national estimates as recommended by the NHIS analytic guidelines. Using SAS 9.4 (PROC SURVEY), we accounted for the complex NHIS survey design and nonresponses using survey weights, sampling unit clusters, and strata. We first analyzed changes in overall CT colonography utilization between 2015 and 2018, and given significant temporal change, we further examined the change for three separate subgroups: by age, sex, and race/ethnicity. We used the same approach to examine the changes in screening CT colonography when restricting to those who were up-to-date on colorectal cancer screening. In this way, we were able to identify those who elected CT colonography for colorectal cancer screening within the recommended period (within 5 years) and to examine whether the individual preference for screening CT colonography followed the overall trend or new user influx caused the overall temporal change. Characteristics of individuals who used CT colonography between 2015 and 2018 were compared using survey design-adjusted Wald F tests. We used a multivariable logistic regression model to examine the predictors of CT colonography adjusting for other covariates (age, sex, race/ethnicity, marital and employment status, education, family income, and health insurance, and survey year). Finally, we used the same multivariable logistic model to compare the predicted probability of CT colonography by more specified types of insurance for those ages 45–64 (private, Medicaid+other public, and uninsured) and ages >65 (Medicare+private, Medicare only, and Medicare+other public) separately, given the established impact of health insurance coverage (20).

Our final study population included 34,768 individuals representing 129,430,319 U.S. adult population ages 45–85 (mean age: 61.5 years; 52.9% female; 71.5% non-Hispanic White). Overall, 54.5% (95% CI, 53.8–55.3) reported being up-to-date on recommended colorectal cancer screening. Of those, 1.8% (95% CI, 1.6–2.1) reported that they chose CT colonography for their colorectal cancer screening. The characteristics of individuals who had CT colonography between 2015 and 2018 were similar (Table 1). However, individuals ages 45–49 (8.5% vs. 2.8%), other race/ethnic group (6.3% vs. 1.0%), and with low family income (37.9% vs. 27.0%) were more likely to report having CT colonography in 2018.

The utilization of CT colonography increased from 0.79% (95% CI, 0.63–0.96) in 2015 to 1.33% (95% CI, 1.11–1.56) in 2018 (P < 0.001; Fig. 1A). Following the overall trends, there was a significant increase in CT colonography use in selected subgroups. For example, the utilization increased across age subgroups ranging from 0.46 to 0.75 percent points for those ages <76+ years old (P < 0.05 for all; Fig. 1B). For race/ethnic groups, a statistically significant increase was observed for non-Hispanic White and other races (0.38 and 1.17 percent points increase, respectively; P < 0.05 for all). Although there was no significant change over time, non-Hispanic Black and Hispanic groups had higher utilization of CT colonography than other race/ethnic groups (Fig. 1D).

Among those being up-to-date on colorectal cancer screening, similar trends were observed between 2015 and 2018 (0.86 percent point increase, P < 0.001; Fig. 2A). The use of CT colonography for colorectal cancer screening increased markedly by 2.59 percent point for those ages 45–49, followed by 0.70 percent point for those ages 65–75 and 0.66 percent point for ages 50–64 (P < 0.05 for all; Fig. 2B). Statistically significant increase in screening CT colonography between 2015 and 2018 was also seen for non-Hispanic White (0.62 percent point increase, P = 0.02) and other race/ethnic groups (2.11 percent point increase, P = 0.01; Fig. 2D). 33.9% (95% CI, 26.3–41.6) of those who had CT colonography for colorectal cancer screening underwent follow-up colonoscopy within the same survey year.

In a fully adjusted logistic regression model (Table 2), increasing age appears to be associated with a lower likelihood of having CT colonography for colorectal cancer screening. For example, those ages 45–49 years were 2.08 times (reversed OR, 2.08; 95% CI, 1.01–4.35) more likely to use CT colonography for colorectal cancer screening than those ages 65+ years. Compared with non-Hispanic White, non-Hispanic Black (OR, 1.83; 95% CI, 1.14–2.93) and Hispanic (OR, 1.84; 95% CI, 1.12–3.03) had a higher likelihood of undergoing screening CT colonography. Other significant predictors of having screening CT colonography included family income <200% of FPL (reversed OR, 1.64; 95% CI, 1.04–2.56) and having publicly funded health insurance (OR, 1.79; 95% CI, 1.04–3.06). The likelihood was no different by sex, marital status, employment, or education.

We estimated and compared the predicted probability of having CT colonography by the further specified type of insurance for those ages 45–64 and 65+ years (Fig. 3). A statistically significant increase between 2015 and 2018 was observed only for those ages 45–64 with Medicare + other public insurance types (P < 0.001). While there was an increase among those 65+ with Medicare only, it did not reach the statistical significance (P = 0.07). Among those 45–64 years of age, the predicted estimate was significantly higher for those with Medicaid + other public (2.47%) compared with private (0.93%, P < 0.001) or uninsured (0.74%, P < 0.001). Among those ages 65+ years, there was no significant difference in predicted estimates of CT colonography by type of supplemental insurance (P = 0.94).

We conducted this study to examine the utilization of CT colonography among a nationally representative sample of U.S. population ages 45 or older. Although the overall utilization of CT colonography was still low (1.3%), we found that there was an increasing proportion of U.S. adults, albeit minor (a 0.86 percent point increase; equivalent to an estimate of 0.25 million Americans), who elected the CT colonography for colorectal cancer screening between 2015 and 2018. This small increase, however, had a significant impact on colorectal cancer screening compliance at the population level. The increase in CT colonography utilization led to an 18.6% increase in overall colorectal cancer screening compliance among U.S. adults ages 45+ years during the study period. We also found that the utilization of CT colonography was higher, especially among those ages 45–49, other race/ethnicity, with low income, and having publicly funded insurance type. Taken together, these findings suggest that CT colonography may be gaining popularity in recent years, mainly among younger adults and socioeconomically disadvantaged groups.

Our findings suggest that CT colonography may be increasing in popularity among younger individuals (ages 45–49), which has important implications for recent changes to ACS screening guidelines. In May 2018, ACS updated its colorectal cancer screening guidelines and lowered the age to initiate screening from 50 to 45 years. Fedewa and colleagues (28) found that the overall colorectal cancer screening rate among individuals ages 45–49 years sharply increased by a 6.9 percent point after the ACS guidelines update. Our findings suggest a similar trend and extend this prior research by demonstrating that increased use of CT colonography in this population may have contributed to the overall increase in colorectal cancer screening uptake in the young adult group. There was a 2.59 percent point increase in screening CT colonography among those ages 45–49 in 2018 in this study. One potential reason for this could be better patient acceptability. Patient acceptability of CT colonography is generally higher than colonoscopy, given that it is less invasive, has a shorter preparation time, and virtually no recovery time (e.g., no sedation) compared with a colonoscopy (29–31). These features may be appealing to those younger age groups who did not wish to undergo colonoscopy. Further studies, especially involving qualitative approach, are needed better to understand younger individuals' preference for colorectal cancer screening.

Our study also found that racial/ethnic minorities (e.g., Black race, Hispanic ethnicity) and individuals with a lower household income (<200% FPL) were more likely to use CT colonography. Disparities in colorectal cancer screening persist by race/ethnicity and socioeconomic status (32–35). In this study, however, the receipt of CT colonography was higher among those racial/ethnic minorities or those with low income, suggesting that these individuals may prefer having CT colonography to optical colonoscopy. Better patient acceptability of CT colonography procedure may also explain this finding (29–31). Besides, for individuals with low-income status, the cost may be another reason. Sawhney and colleagues (36) estimated that the average screening colonoscopy cost was $1,425 (2016 USD), whereas the average cost for CT colonography was $1,081 among a commercially insured population. In Medicare study (37), the estimated average CT colonography cost was $439; however, the average colonoscopy cost was $1,035 per procedure (2015 USD). Overall, having CT colonography provides a lower cost (24%–57% less) option, along with the other advantages. Therefore, it is possible that promoting CT colonography could be an effective strategy to improve colorectal cancer screening among socioeconomically disadvantaged populations. Considering established disparities in health literacy and patient comprehension of cancer screening in these populations (38–40), assessing patient acceptance and perceived affordability of CT colonography compared with other cost-effective options (e.g., FOBTs) would be an important avenue for future research.

Our findings could have important implications for health policymakers and insurers. Having insurance coverage is one of the strongest predictors of cancer screening and CT colonography use (20). We also observed that insurance coverage, specifically public type for those ages 45–64 years, was associated with higher utilization of CT colonography. State-mandated colorectal cancer screening coverages (e.g., California, Maryland, Delaware) including state Medicaid cover any colorectal cancer tests recommended by USPSTF (41, 42). This may be one of the factors contributing to higher utilization of CT colonography among individuals ages 45–64 years with public and state insurance programs. The pattern for Medicare is different; however, we did not observe any distinctive difference in CT colonography use among those 65+ years of age. One of the major factors is that the Centers for Medicare and Medicaid Services (CMS)—the biggest payer in the U.S. healthcare system—does not cover CT colonography as a primary colorectal cancer screening test for Medicare beneficiaries (43). Despite the recent endorsement of CT colonography as an acceptable method for colorectal cancer screening and its inclusion in the guidelines by USPSTF in 2016 (3), CMS does not provide Medicare coverage for CT colonography. Under the ACA, many private insurers and state-based programs already covered CT colonography with no cost sharing (20, 44). Given the higher colorectal cancer prevalence in the elderly population (1, 2), Medicare coverage of CT colonography could improve colorectal cancer screening by providing a less invasive alternative to a colonoscopy. Despite the potential of CT colonography to improve screening disparities while addressing the limitations of colonoscopy, effective implementation of CTC for colorectal cancer screening may be challenging under the current CMS reimbursement policies (20, 45). Future studies should examine the long-term economic and health benefits of CT colonography utilization in the Medicare population to better inform U.S. healthcare policy decisions.

There are limitations to this study and interpretation of the results needs some caution. First, we relied on the self-reported data including CT colonography use, which could be subject to reporting bias and overestimation. However, past validation studies have demonstrated comparable estimates between self-reported information and medical records for cancer screening (46, 47). Second, the nature of secondary data analysis did not allow us to specify the reasons for CT colonography use and follow-up care (e.g., for diagnostic tests or other health problems). Because CT colonography scans the entire abdominal cavity, it could detect extracolonic findings (11). Studies reported that extracolonic findings were common during CT colonography and occurred 23%–66% of those who underwent CT colonography examination (12, 29). Future studies are needed to define and quantify benefits and harms associated with follow-up care after CT colonography. Third, our study was not able to capture some of the modifiable factors that influence cancer screening uptakes, such as health beliefs or provider recommendation (39, 48). Fourth, our study was only able to capture initial colorectal cancer screening, not repeat screening. Repeat screening for other colorectal cancer screening methods remains low and additional research is needed to determine whether using CT colonography, which may have advantages over colonoscopy, improves repeat screening (49, 50).

Our study suggests that the use of CT colonography is increasing, but the overall reach remains low. Younger individuals, racial and ethnic minorities, or those with low income appear to have a higher CT colonography utilization. CT colonography is a promising screening modality for colorectal cancer. To support colorectal cancer prevention efforts, future research is needed to test implementation strategies that increase the usage of CT colonography (e.g., physician recommendation, patient navigation/education programs) and further explore patient awareness/preferences in colorectal cancer screening options and potential barriers to using CT colonography.

No disclosures were reported.

YR. Hong: Conceptualization, resources, data curation, software, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing-original draft, project administration, writing-review and editing. Z. Xie: Data curation, software, formal analysis, methodology. K. Turner: Conceptualization, data curation, formal analysis, validation, investigation, methodology, writing-original draft, project administration, writing-review and editing. S. Datta: Validation, investigation, writing-review and editing. R. Bishnoi: Conceptualization, data curation, validation, investigation, methodology, writing-original draft, project administration, writing-review and editing. C. Shah: Conceptualization, validation, investigation, methodology, writing-original draft, writing-review and editing.

The authors gratefully acknowledge support from the Department of Health Services Research, Management and Policy in the College of Public Health and Health Professions at the University of Florida.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.