Abstract
Background and Aims
At the onset of the coronavirus disease 2019 (COVID-19) pandemic, all elective outpatient procedures were halted by Ohio public health authorities between March 2020 and June 2020, leading to limited screening colonoscopies. In this study, we studied the impact of COVID-19 on trends and disparities on screening colonoscopy use.
Methods
We identified all patients who underwent colonoscopy at Cleveland Clinic Ohio facilities in 2019, 2020, and 2021 from July 1 to December 31 in each year. We then calculated rates of screening colonoscopy and factors associated with colonoscopy use during the study periods.
Results
Among 68,206 colonoscopies, 23,316 (34.2%) were screening colonoscopies. In whites, the rate of colonoscopy use increased in 2021 (76.9%) over 2019 (46%) levels, whereas use of colonoscopy in African Americans decreased to even less than 2019 rates (16.5%) in 2021 (15.9%) (P < .001). Patients with private insurance and the highest quartiles of median household income and education had increasing rates of colonoscopy (P < .001) in 2020 and 2021. Patients who had ≥2 polyps and adenomas significantly increased from 2019 to 2021 (P < .05 for both). On multivariate analysis, age >65 years, male gender, white race, tobacco use, private insurance, and obesity were positive predictors of adenomas (P < .05 for all).
Conclusions
The pandemic exacerbated pre-existing healthcare disparities in colonoscopy use, which have continued to persist in 2021. These data will assist current and future efforts to increase uptake of colorectal cancer screening, especially in historically marginalized ethnic and low socioeconomic communities.
Colorectal cancer (CRC) screening is the cornerstone of CRC prevention and is recommended for all average-risk adults at age 45 years in the United States. It has resulted in 35% and 37% reduction in CRC incidence and mortality, respectively.1 Previous studies have reported lower rates of CRC screening in racial and ethnic minorities and in persons with inequities in socioeconomic factors such as lower education level and household income.2 At the onset of the coronavirus disease 2019 (COVID-19) pandemic, all elective outpatient procedures were halted by Ohio public health authorities between March 2020 and June 2020, leading to limited screening colonoscopies. COVID-19 has exacerbated these pre-existing healthcare disparities, raising concerns of missed opportunities and decreased CRC screening in groups that have been economically and/or socially marginalized.3, 4, 5, 6
Several studies reported significantly decreased CRC screening during the pandemic, with most reporting a >50% decrease in endoscopy volume.7, 8, 9 However, although these studies have looked at healthcare access disparities during the first surge of COVID-19, they have not studied whether these trends in healthcare access have persisted in 2021 after the country had adequate time to become accustomed to the pandemic landscape. Therefore, the true gravity of these healthcare access disparities on CRC screening has not been reported in the post–COVID-19 lockdown era.
In this study, we studied the impact of the COVID-19 pandemic on trends and demographic and socioeconomic disparities in screening colonoscopy use. We also studied the demographic and socioeconomic factors associated with adenoma detection rates (ADRs).
Methods
After obtaining institutional review board approval from the Cleveland Clinic, we identified all patients who underwent colonoscopy at the Cleveland Clinic Ohio facilities in 2019, 2020, and 2021 from July 1 to December 31 in each year. All Cleveland Clinic Ohio facilities, including tertiary care centers, hospitals, and outpatient endoscopy centers, were included. The timeline of July 1 to December 31 was chosen because all elective outpatient procedures were halted by Ohio public health authorities between March 2020 and June 2020, leading to limited screening colonoscopies, and because we sought to compare factors before, immediately after, and a year after COVID-19 lockdown to understand its long-term impact.
The following data were extracted by using electronic medical records by natural language processing algorithms: age, sex, race, socioeconomic status, zip code of residence, employment status, marital status, primary language, type of insurance (Medicare, Medicaid, other public insurance, commercial, or uninsured), indication of examination (screening, surveillance, or diagnostic) substance abuse (tobacco, alcohol, or illicit drugs), body mass index (BMI), and location, size, and number of polyps found on colonoscopy. We then calculated rates of screening colonoscopy and factors associated with colonoscopy use during the study periods.
Data for household income were not available from electronic medical records. Therefore, as a proxy, median household income for patients’ residential addresses and percentage of high school graduates in the zip code were obtained from U.S. Census American Community Survey data 2010 estimates.10 Median household income and education were then divided into quartiles based on the distribution, which is used as a marker of socioeconomic status.
In addition, data on pathology including adenomas or malignancy were extracted from pathology reports. Patients were categorized with an adenoma pathology if they had villous, tubular, or tubulovillous adenoma histology on any polyps. Advanced adenoma was defined as polyps ≥10 mm or advanced histology (villous or tubulovillous) or presence of high-grade dysplasia. We excluded patients with serrated lesions and hyperplastic polyps. The patients were divided into 2 groups: patients with an adenoma or other precancerous polyps (cases) and patients without an adenoma (control subjects) detected on colonoscopy.
Statistical analysis
Baseline demographics and outcomes of interest were compared between both groups. Categorical variables are expressed as percentages and were compared by the Pearson χ2 method. Continuous variables are expressed as median (interquartile range) or mean ± standard deviation and were compared by a Kruskal-Wallis test and an independent t test, respectively. Two-sided P values were reported with a significance level of .05. All data were analyzed using SPSS version 25.0 (IBM corporation, Armonk, NY, USA). Any variable with an alpha ≤ .05 was included in the logistic regression model for ADR. Odds ratio was calculated to estimate the predictors of ADR.
Results
Of 68,206 colonoscopies performed during the study period, 23,316 (34.2%), 18,931 (27.8%), and 25,959 (38.1%) were screening, surveillance, and diagnostic colonoscopies, respectively. The proportion of screening colonoscopies significantly declined in 2020 (31.2%) and then increased to 37.2% in 2021 as compared with 34.8% in 2019 (P < .001) (Fig. 1). The mean age of patients undergoing screening colonoscopy significantly went down after the COVID-19 pandemic (P < .001), but the proportion of seniors (age >65 years) and the male-to-female ratio remained the same. In whites, the rate of colonoscopy use initially declined in 2020 (74.2%) and then increased in 2021 (76.9%) over 2019 (46%) levels, whereas use of colonoscopy in African Americans initially increased in 2020 (18.9%) and then decreased to even lower than 2019 rates (16.5%) in 2021 (15.9%) (P < .001) (Table 1). Patients with private insurance and the highest quartiles of median household income and education had increasing rates of screening colonoscopies (P < .001) in 2020 and 2021. Patients who had ≥2 polyps (12.2%, 13.6%, and 15.1%, respectively, for 2019, 2020, and 2021), adenomas (21.5%, 22.3%, and 23.7%, respectively), and polyps in the right-sided colon (45.5%, 52.4%, and 53.4%, respectively) significantly increased from 2019 to 2021 (P < .05 for all) (Table 1 and Fig. 1). Fortunately, no significant increase was noted in advanced ADRs (4.6% vs 4.8% vs 4.4%, P = .48) or CRC (.2% vs .2% vs .1%, P = .45) from 2019 to 2021 (Fig. 1).
Figure 1.
Trends of screening colonoscopy, ≥2 polyps, adenomas, advanced adenomas, and new diagnosis of CRC in 2019, 2020, and 2021. CRC, Colorectal cancer.
Table 1.
Demographics, socioeconomic, and colonoscopy findings during the study period
| Factor | 2019 (n = 7905) | 2020 (n = 6737) | 2021 (n = 8674) | P value |
|---|---|---|---|---|
| Age, y | 59.8 ± 8.6 | 59.6 ± 8.9 | 58.9 ± 9.3 | <.001 |
| Age ≥65 years | 2275 (28.8) | 1871 (27.8) | 2472 (28.5) | .38 |
| Sex | .97 | |||
| Male | 3648 (46.1) | 3107 (46.1) | 4014 (46.3) | |
| Female | 4257 (53.9) | 3630 (53.9) | 4660 (53.7) | |
| Race | <.001 | |||
| White | 6004 (76) | 4999 (74.2) | 6672 (76.9) | |
| African American | 1303 (16.5) | 1271 (18.9) | 1381 (15.9) | |
| Other | 598 (7.6) | 467 (6.9) | 621 (7.2) | |
| Insurance type | <.001 | |||
| Medicare | 1965 (24.9) | 1619 (24) | 1868 (21.5) | |
| Medicaid and other public | 435 (5.5) | 491 (7.3) | 541 (6.2) | |
| Private | 4662 (59) | 4061 (60.3) | 5641 (65) | |
| No insurance | 843 (10.7) | 566 (8.4) | 624 (7.2) | |
| Education level (based on percentage of high school graduates in the zip code) | <.001 | |||
| Quartile 1: <88 | 1905 (22.8) | 1557 (24.1) | 1768 (21.3) | |
| Quartile 2: ≥88 to <92.5 | 1815 (24.3) | 1503 (23.3) | 1923 (23.1) | |
| Quartile 3: ≥92.5 to <94 | 1714 (22.9) | 1549 (22.6) | 1942 (23.3) | |
| Quartile 4: ≥94 | 2237 (29.9) | 1935 (30) | 2686 (32.3) | |
| Median household income | <.001 | |||
| Quartile 1: <43,449 | 1647 (22) | 1544 (23.9) | 1715 (20.6) | |
| Quartile 2: ≥43,449 to <55,969 | 1837 (24.6) | 1532 (23.7) | 1964 (23.6) | |
| Quartile 3: ≥55,969 to <67,917 | 1889 (25.3) | 1569 (24.3) | 2154 (25.9) | |
| Quartile 4: ≥67,917 | 2098 (28.1) | 1809 (28) | 2486 (29.9) | |
| Tobacco use | 699 (8.8) | 636 (9.4) | 827 (9.5) | .13 |
| Alcohol use | 5033 (63.7) | 4310 (64) | 5747 (66.3) | .001 |
| Illicit drug use | 259 (3.3) | 245 (3.6) | 326 (3.8) | .22 |
| Preferred language | .5 | |||
| English | 7732 (97.8) | 6613 (98.2) | 8489 (97.9) | |
| Spanish | 83 (1) | 60 (.9) | 81 (.9) | |
| Other | 90 (1.1) | 64 (.9) | 104 (1.2) | |
| Body mass index, kg/m2 | 29.6 ± 6.4 | 29.8 ± 6.5 | 29.7 ± 6.5 | .52 |
| Colonoscopy findings | ||||
| No. of polyps | <.001 | |||
| 0 | 4483 (56.7) | 3770 (56) | 4643 (53.5) | |
| 1 | 2456 (31.1) | 2049 (30.4) | 2721 (31.4) | |
| ≥2 | 966 (12.2) | 918 (13.6) | 1310 (15.1) | |
| Adenoma | 1697 (21.5) | 1503 (22.3) | 2059 (23.7) | .002 |
| Advanced adenoma | 361 (4.6) | 324 (4.8) | 382 (4.4) | .48 |
| Colorectal cancer | 12 (.2) | 12 (.2) | 9 (.1%) | .45 |
Values are mean ± standard deviation or n (%).
As compared with control subjects, a significantly higher number of older patients (mean age, 60.6 ± 9.0 years vs 59.1 ± 8.9 years; P < .001), men (55.1% vs 43.6%, P < .001), and whites (77.1% vs 75.4%, P = .003) were found among the cases. Patients with private insurance (58.3% vs 62.6%) had significantly decreased ADRs, whereas patients with Medicare (27% vs 22.3%) and Medicaid (6.8% vs 6.2%) had higher ADRs as compared with control subjects. Significantly fewer patients were in the highest quartiles of education (29.3% vs 31.3%) and median household income (26.8% vs 29.3%) among cases than control subjects (P < .05 for both). Patients who used tobacco and with a high BMI also had significantly higher ADRs (P < .001), whereas alcohol use, illicit drug use, and preferred language were not significantly associated with higher ADRs (P > .05 for all). On multivariate analysis, age >65 years, male gender, white race, tobacco use, private insurance, and high BMI were positive predictors of adenomas (P < .05 for all) (Table 2).
Table 2.
Demographic and socioeconomic factors associated with adenoma detection on screening colonoscopy
| Factors | No adenoma (n = 18,057) | Adenoma (n = 5259) | P value | Adjusted odds ratio (95% confidence interval) | P value |
|---|---|---|---|---|---|
| Age, y | 59.1 ± 8.9 | 60.6 ± 9.0 | <.001 | NA | |
| Age ≥65 y | 4877 (27) | 1741 (33.1) | <.001 | 1.7 (1.6-1.7) | <.001 |
| Sex | |||||
| Male | 7873 (43.6) | 2896 (55.1) | <.001 | Reference | |
| Female | 10,184 (56.4) | 2363 (44.9) | .65 (.62-.67) | <.001 | |
| Race | .003 | ||||
| Whites | 13,621 (75.4) | 4054 (77.1) | Reference | ||
| African American | 3145 (17.4) | 810 (15.4) | .9 (.85-.96) | .001 | |
| Other | 1291 (7.1) | 395 (7.5) | .97 (.89-1.05) | .54 | |
| Insurance type | <.001 | ||||
| Medicare | 4033 (22.3) | 1419 (27) | 1.04 (.98-1.09) | .15 | |
| Medicaid and other public | 1111 (6.2) | 356 (6.8) | .87 (.8-.94) | .001 | |
| Private | 11,297 (62.6) | 3067 (58.3) | Reference | ||
| No insurance | 1616 (8.9) | 417 (7.9) | .76 (.71-.82) | <.001 | |
| Education level based on percentage of high school graduates in the zip code) | .008 | ||||
| Quartile 1: <88 | 3887 (22.6) | 1143 (22.7) | Reference | ||
| Quartile 2: ≥88 to <92.5 | 3973 (23.1) | 1268 (25.1) | 1.01.(.94-1.08) | .79 | |
| Quartile 3: ≥92.5 to <94 | 3962 (23) | 1153 (22.9) | .99 (.91-1.08) | .82 | |
| Quartile 4: ≥94 | 5380 (31.3) | 1478 (29.3) | .99 (.9-1.09) | .87 | |
| Median household income (quartiles based on zip code) | .005 | ||||
| Quartile 1: <43,449 | 3759 (21.9) | 1147 (22.7) | Reference | ||
| Quartile 2: ≥43,449 to <55,969 | 4113 (23.9) | 1220 (24.2) | .96 (.90-1.03) | .37 | |
| Quartile 3: ≥55,969 to <67,917 | 4288 (24.9) | 1324 (26.3) | 1.08 (.98-1.18) | .09 | |
| Quartile 4: ≥67,917 | 5042 (29.3) | 1351 (26.8) | 1.0 (.9-1.1) | .99 | |
| Tobacco use | 1549 (8.6) | 613 (11.7) | <.001 | 1.41 (1.32-1.51) | <.001 |
| Alcohol use | 11,686 (64.7) | 3404 (64.7) | .99 | NA | |
| Illicit drug use | 622 (3.4) | 208 (4) | .079 | NA | |
| Preferred language | .3 | NA | |||
| English | 17,690 (98) | 5144 (97.8) | |||
| Spanish | 176 (1) | 48 (0.9) | |||
| Other | 191 (1.1) | 67 (1.3) | |||
| Body mass index, kg/m2 | 29.5 ± 6.5 | 30.3 ± 6.5 | <.001 | ||
| Obesity | 7200 (40.3) | 2333 (44.8) | <.001 | 1.28 (1.23-1.33) | <.001 |
Values are mean ± standard deviation or n (%). NA, not applicable.
Discussion
In this large study of screening colonoscopies, we found that the COVID-19 pandemic exacerbated pre-existing healthcare disparities in colonoscopy use that have continued to persist in 2021. In addition, the number of polyps and adenomas and polyp location detected on screening colonoscopy have significantly increased in the postpandemic lockdown period, but that did not translate into excess advanced ADRs or CRC detection rates.
Several studies have reported decreased rates of screening colonoscopies and CRC detection associated with the pandemic.7, 8, 9,11 Our findings explore and provide much needed data on the impact of the pandemic in further exaggerating healthcare disparities in screening colonoscopies. In addition, we not only studied the impact of the pandemic immediately after the initial lockdown period, where one expects decreased numbers, but reported data 1 year after the initial lockdown. The rate of screening colonoscopies after the pandemic in communities of color (African Americans) initially increased in 2020 but then decreased in 2021, whereas patients with high education and income and private insurance had a significantly increased rate as compared with 2019. This is a concerning trend because higher education and income and private insurance typically translate into higher socioeconomic status and health literacy. Similar to prior studies, the pandemic disproportionately affected communities of color and historically marginalized populations possibly because of systemic racism, presence of multiple comorbidities, access to health services, and other socioeconomic factors, leading to lower rates of CRC screening.4,5 Pre-existing systemic-based barriers such as loss of job with a lapse in insurance, access to specialists, and financial barriers including indirect costs (parking and need for a driver) exacerbated pre-existing racial disparities in the use of screening colonoscopy.12 New protocols for outpatient procedures such as preprocedure testing and a limited ability to do virtual visits (access to mobile, internet, or smart devices), leading to even more decreased contact with the healthcare system, likely also contributed to these findings.13
Our study also highlights that the number of polyps and adenomas detected on screening colonoscopy have significantly increased in the postpandemic period. In addition, more adenomas were detected in the right-sided colon (proximal to splenic flexure) as compared with prepandemic procedures. Fortunately, we did not find increasing rates of advanced adenomas or CRC. A decrease in CRC detection because of the pandemic has been reported in studies from Europe, Asia, and the United States.7, 8, 9,11,14 Data from the GIQuIC Registry reported 30% fewer CRC diagnoses than expected in the first 7 months of the pandemic as compared with prepandemic values, similar to other studies.7 Because our study included only screening colonoscopies and excluded surveillance or diagnostic examinations along with data from the postpandemic lockdown period are likely reasons for the contrasting finding. Nevertheless, a higher number of adenomas and no difference in CRC makes biologic plausible sense as the adenoma to carcinoma sequence typically progresses slowly.15 In 2020, an increase in stool testing such as with fecal immunochemical tests (7%) and a decrease in colonoscopy (16%) were reported.7,16 Our study also found that age >65 years, male gender, white race, tobacco use, private insurance, and a high BMI were positive predictors of adenomas on screening colonoscopies (P < .05 for all). Timely resumption of screening efforts along with community outreach to historically marginalized populations are needed to maintain the down-trending incidence of CRC attributed to early detection. The increasing ADR trend could potentially lead to a rising incidence of CRC in the future if efforts to increase CRC screening uptake are not successful. In addition, particular efforts in minimizing delay in colonoscopy after positive stool tests are paramount in preventing excess CRC rates and mortality.6
Our study has several strengths and limitations. Our strengths are a large sample size and generalizable data encompassing a large geographic area including tertiary care, hospital-based, and outpatient endoscopy centers. In addition, data from pathology along with a 3-year trend also add to the strength of the study. Limitations are that data were extracted from electronic medical records using a natural language processing algorithm, single-state data were used, lack of other racial and ethnic populations other than black communities, lack of data on surveillance or diagnostic colonoscopies and other methods of CRC screening such as fecal immunochemical tests, and lack of data on serrated lesions. This retrospective cohort study used patient zip codes to determine median household income and education level based on census data. Although this method of determining socioeconomic status is efficient, there is some room for error because of the great variability that can exist in a zip code. Nevertheless, this is the largest study to report on the impact of the COVID-19 pandemic on exacerbating healthcare disparities on screening colonoscopy.
In conclusion, the pandemic exacerbated pre-existing racial and socioeconomic healthcare disparities in screening colonoscopy use, which have continued to persist in 2021. The number of polyps and adenomas detected on screening colonoscopy has significantly increased in the postpandemic lockdown period, but that did not translate into excess advanced ADRs or CRC rates. These findings signify the need for increased measures to improve the uptake of CRC screening, especially in historically marginalized ethnic and low socioeconomic communities.
Disclosure
All authors disclosed no financial relationships.
References
- 1.Issa I.A., Noureddine M. Colorectal cancer screening: an updated review of the available options. World J Gastroenterol. 2017;23:5086–5096. doi: 10.3748/wjg.v23.i28.5086. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Joseph D.A., Redwood D., DeGroff A., et al. Use of evidence-based interventions to address disparities in colorectal cancer screening. MMWR Morb Mortal Wkly Rep. 2016;65(Suppl):21–28. doi: 10.15585/mmwr.su6501a5. [DOI] [PubMed] [Google Scholar]
- 3.Dorn A.V., Cooney R.E., Sabin M.L. COVID-19 exacerbating inequalities in the US. Lancet. 2020;395:1243–1244. doi: 10.1016/S0140-6736(20)30893-X. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Ruprecht M.M., Wang X., Johnson A.K., et al. Evidence of social and structural COVID-19 disparities by sexual orientation, gender identity, and race/ethnicity in an urban environment. J Urban Health. 2021;98:27–40. doi: 10.1007/s11524-020-00497-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Tai D.B.G., Shah A., Doubeni C.A., et al. The disproportionate impact of COVID-19 on racial and ethnic minorities in the United States. Clin Infect Dis. 2021;72:703–706. doi: 10.1093/cid/ciaa815. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Balzora S., Issaka R.B., Anyane-Yeboa A., et al. Impact of COVID-19 on colorectal cancer disparities and the way forward. Gastrointest Endosc. 2020;92:946–950. doi: 10.1016/j.gie.2020.06.042. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Calderwood A.H., Calderwood M.S., Williams J.L., et al. Impact of the COVID-19 pandemic on utilization of EGD and colonoscopy in the United States: an analysis of the GIQuIC Registry. Tech Innov Gastrointest Endosc. 2021;23:313–321. doi: 10.1016/j.tige.2021.07.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Khan A., Bilal M., Morrow V., et al. Impact of the coronavirus disease 2019 pandemic on gastrointestinal procedures and cancers in the United States: a multicenter research network study. Gastroenterology. 2021;160:2602–2604. doi: 10.1053/j.gastro.2021.02.055. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Lee J.K., Lam A.Y., Jensen C.D., et al. Impact of the COVID-19 pandemic on fecal immunochemical testing, colonoscopy services, and colorectal neoplasia detection in a large United States community-based population. Gastroenterology. 2022;163:723–731. doi: 10.1053/j.gastro.2022.05.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.U.S. Census Bureau American Community Survey Data. Published November 23, 2021. https://www.census.gov/programs-surveys/acs/data.html Available at:
- 11.Castonguay M., El Sayed R., Richard C., et al. COVID-19 impact on diagnosis and staging of colorectal cancer: a single tertiary canadian oncology center experience. Curr Oncol. 2022;29:3282–3290. doi: 10.3390/curroncol29050268. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.McCarty T.R., Hathorn K.E., Bazarbashi A.N., et al. Endoscopic gallbladder drainage for symptomatic gallbladder disease: a cumulative systematic review meta-analysis. Surg Endosc. 2021;35:4964–4985. doi: 10.1007/s00464-020-07758-3. [DOI] [PubMed] [Google Scholar]
- 13.Eberly L.A., Kallan M.J., Julien H.M., et al. Patient characteristics associated with telemedicine access for primary and specialty ambulatory care during the COVID-19 pandemic. JAMA Netw Open. 2020;3 doi: 10.1001/jamanetworkopen.2020.31640. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Morris E.J.A., Goldacre R., Spata E., et al. Impact of the COVID-19 pandemic on the detection and management of colorectal cancer in England: a population-based study. Lancet Gastroenterol Hepatol. 2021;6:199–208. doi: 10.1016/S2468-1253(21)00005-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Patel S.G., May F.P., Anderson J.C., et al. Updates on age to start and stop colorectal cancer screening: recommendations from the U.S. Multi-Society Task Force on Colorectal Cancer. Gastroenterology. 2022;162:285–299. doi: 10.1053/j.gastro.2021.10.007. [DOI] [PubMed] [Google Scholar]
- 16.Fedewa S.A., Star J., Bandi P., et al. Changes in cancer screening in the US during the COVID-19 pandemic. JAMA Netw Open. 2022;5 doi: 10.1001/jamanetworkopen.2022.15490. [DOI] [PMC free article] [PubMed] [Google Scholar]