Secondary Logo

Journal Logo

Epidemiology

Colorectal Cancer Screening in People With and Without HIV in an Integrated Health Care Setting

Lam, Jennifer O. PhD, MPHa; Hurley, Leo B. MPHa; Udaltsova, Natalia PhDa; Alexeeff, Stacey E. PhDa; Klein, Daniel B. MDb; Corley, Douglas A. MD, PhDa; Silverberg, Michael J. PhD, MPHa

Author Information
JAIDS Journal of Acquired Immune Deficiency Syndromes: July 1, 2019 - Volume 81 - Issue 3 - p 284-291
doi: 10.1097/QAI.0000000000002024

Abstract

INTRODUCTION

As effective antiretroviral treatment has reduced the burden of AIDS-defining malignancies in people with HIV (PWH), malignancies not traditionally associated with HIV/AIDS now account for an increasing proportion of cancers.1 Screening for cancers that are common in the general population is an increasingly important component of care, especially as PWH experience longer life expectancy and reach ages when cancer screening would typically be recommended.

Colorectal cancer (CRC) is the third most common nonskin cancer and the second leading cause of cancer deaths in the United States.2 As CRC is relatively slow-progressing, detecting and removing precancerous adenomatous polyps (adenomas) is a highly effective method of preventing CRC, which has led to declining rates of CRC cases and deaths in the general population.3,4 Most guidelines recommend screening of asymptomatic persons at average risk for CRC starting at age 50, with earlier screening initiation and more frequent follow-up for patients at higher risk (ie, family history of CRC or personal medical condition that increases risk for CRC, such as polyposis syndrome).5

Evidence on the incidence of CRC in PWH relative to persons without HIV has been inconsistent. A meta-analysis that pooled estimates from 27 studies of 23 populations through June 2016 concluded that CRC incidence in PWH and uninfected populations is similar [standardized incidence ratio: 1.00, 95% confidence interval (CI) 0.82 to 1.22].6 However, heterogeneity among the studies being compared was significant (I2 = 89.2%) and the studies did not consistently account for important individual-level factors such as smoking, body mass index (BMI), genetic predisposition to CRC (ie, family history), and severity of HIV disease.6 In contrast to the conclusions of the meta-analysis, recent results from the HIV/AIDS Cancer Match Study, the largest study of cancer in PWH in the United States, indicate that rates of CRC are approximately 30%–50% lower in PWH compared with the general population,7,8 depending on anatomical site (ie, proximal or distal colon, or rectum), although this lower rate may be explained by the exclusion of rectal tumors with squamous histology, which are more common in PWH and could potentially be misclassified anal cancers.9 Recent studies within our health system found that although there was no overall difference in CRC incidence in PWH compared with persons without HIV,10 PWH were diagnosed with CRC at a younger mean age (53 vs. 58 years) and at a similar cancer stage,10 and 1.8-fold higher CRC incidence was observed among patients with more severe HIV infection (ie, CD4 <200 cells/µL).11 Other studies have reported that PWH diagnosed with CRC have less favorable outcomes and increased mortality, possibly because CRC has a more aggressive clinical progression in the HIV setting.12–16 Therefore, it remains unclear whether CRC screening guidelines intended for the general population would be appropriate for PWH, particularly those with advanced HIV disease, or whether a different approach such as earlier screening would be required to achieve clinically impactful CRC prevention in this population.

Currently, little is known about CRC screening practices and outcomes in PWH.17–19 Several studies have reported higher adenoma detection in PWH,6,16,20–22 which would ordinarily be associated with higher cancer risk. However, few studies have examined CRC screening outcomes, while also accounting for screening rates in a comparable control population with similar access to care and the ones that have were limited by small sample sizes.16,21 Studies have also not consistently reported or accounted for CRC stage at diagnosis.6

The primary aim of this study was to look for any differences in PWH compared with persons without HIV across several facets of CRC screening, including screening initiation, modality, and outcomes in a large insured population in the United States. The secondary aim was to determine factors associated with detection of adenoma or CRC in PWH.

METHODS

Study Design, Setting, and Population

We conducted an observational cohort study of PWH and demographically matched persons without HIV at Kaiser Permanente Northern California (KPNC) between 2005 and 2016. KPNC is a large integrated health care network, providing comprehensive health care to a diverse patient population demographically representative of insured Northern Californians.23 KPNC promotes CRC screening in accordance with national guidelines to all enrolled members, using a mailed fecal immunochemical testing (FIT)-based outreach program and physician-ordered screening colonoscopy.24 PWH were identified using KPNC's internal HIV registry, which maintains a database of all known cases by monitoring International Classification of Diseases (ICD)–coded patient encounters and laboratory and pharmacy databases for indicators of probable HIV infection, and verifying potential cases by review of medical charts. We included KPNC patients who were aged 50–75 years at any time during the study period, had at least 12 months of health plan membership before start of follow-up (to allow assessment of CRC- and HIV-related medical history), and had no known previous CRC screening by FIT, colonoscopy, or sigmoidoscopy. Follow-up started on the latest of: the patient's 50th birthday, January 1, 2005, or the first day after January 1, 2005 that they met the enrollment criteria. Thus, for some patients, follow-up started after age 50. Follow-up ended on the earliest of: date of outcome of interest, end of study period (December 31, 2016), last day aged 75, last day of continuous health plan membership ignoring gaps of 3 months or less, or death.

To obtain a cohort of patients at average risk of CRC and eligible for CRC screening according to current guidelines,5 we excluded patients with documented CRC screening (including FIT, fecal occult blood test, sigmoidoscopy, or colonoscopy) before age 50; previous adenoma or CRC diagnoses; or, history of inflammatory bowel disease (IBD), colectomy, or polyposis syndrome. ICD codes, version 9 (ICD-9) and version 10 (ICD-10), and procedural codes were used to identify first CRC-related screening and diagnoses, and an algorithm was used to exclude patients who had their first KPNC sigmoidoscopy or colonoscopy performed for clinical indications (ie, for diagnostic or surveillance purposes, or as a follow-up to a previous CRC screening).25 HIV-uninfected comparators meeting the same eligibility criteria were randomly sampled from the KPNC patient population and frequency-matched at a ratio of 10:1 to PWH by age, sex, race/ethnicity, and the medical center where the patient received most of their care, to ensure that the exposed cohort (PWH) and the unexposed cohort (persons without HIV) were both sampled from an underlying population with the same distribution of key demographic covariates.26 The study protocol was reviewed and approved by the KPNC IRB.

Study Measures

CRC screening within KPNC was determined by reviewing patients' electronic health records for CRC-related diagnoses and procedures. CRC screening obtained at non-KPNC facilities was ascertained by reviewing a billing claims database. The date of each patient's first CRC screening (FIT, sigmoidoscopy, or colonoscopy) and whether adenoma or CRC was detected at sigmoidoscopy or colonoscopy were gathered from electronic health records. Data were also gathered on the following covariates: sex, age, race/ethnicity, primary medical center, number of outpatient visits in the year before start of follow-up, smoking status, BMI, history of type-2 diabetes, and IBD diagnosed after start of follow-up. For PWH, we also collected information on duration of HIV infection, nadir CD4 count in KPNC, CD4 count at start of follow-up, HIV RNA level at start of follow-up, use of antiretroviral medications, and previous clinical AIDS diagnosis.

Statistical Analyses

Baseline nonmatched characteristics of PWH and persons without HIV were compared using t tests for continuous variables and χ2 tests for categorical variables. Kaplan–Meier estimates and Cox proportional hazards models were used to evaluate time to receipt of first CRC screening by HIV status, with age as the time scale and estimates expressed as hazard ratios (HRs). Models to calculate adjusted HRs included terms for sex, race/ethnicity, number of outpatient visits, smoking status, BMI, and type-2 diabetes. Among the subset of screened patients, we compared sigmoidoscopy and colonoscopy screening outcomes between HIV and non-HIV groups by calculating crude and adjusted prevalence ratios (PRs) using Poisson regression with robust error variance.27 Models to calculate adjusted PRs included terms for sex, age, race/ethnicity, smoking status, BMI, type-2 diabetes, and IBD diagnosed after start of follow-up.

Matched covariates were included in adjusted models to account for potential confounding by these variables among the subset of screened patients. The effect of each of these terms on screening outcome was also examined separately in PWH and persons without HIV, and then in the overall cohort to test for interaction by HIV status. Among PWH, we evaluated whether HIV-related factors were associated with screening outcome, with the primary predictor being CD4 count (<200/200–499/≥500 cells/µL).

Because some patients entered follow-up after age 50 (ie, enrolled as a KPNC member at age >50 years), we conducted sensitivity analyses among patients who were age 50 at start of follow-up, thus evaluating a subset of the full cohort from the time of recommended screening to the time of first screening. Statistical analyses were conducted using SAS 9.3, Cary, NC, and a 2-tailed P-value <0.05 was considered statistically significant.

RESULTS

Cohort Characteristics

A total of 3177 PWH and 29,219 persons without HIV who were eligible for CRC screening were included in the study. Baseline characteristics of study participants are presented in Table 1. Most study participants were men (91%) and White (59%), and mean age was 53 years. Compared to persons without HIV, PWH had more outpatient visits (average 8 vs. 4 visits, P < 0.001), were more likely to have a history of smoking (48% vs. 35%, P < 0.001), and were less likely to be obese (18% vs. 34%, P < 0.001). Most PWH had been treated with antiretroviral medications (90%).

TABLE 1.
TABLE 1.:
Characteristics of Cohort at Start of Follow-Up, by HIV Status—Kaiser Permanente Northern California, 2005–2016

Time to First CRC Screening, Factors Associated With Screening, and Type of Screening Test

Within 1 year of follow-up, 42% of PWH and 35% of persons without HIV were screened for CRC (P < 0.001; Fig. 1). The cumulative proportion of patients who completed a CRC screen of any type within 5 years was 86% in PWH and 79% in persons without HIV (P < 0.001; Fig. 1). During the first 5 years of follow-up, time to first CRC screening was faster among PWH (HR: 1.23, 95% CI: 1.17 to 1.28; adjusted HR: 1.06, 95% CI: 1.01 to 1.11). The associations of sex, race/ethnicity, number of outpatient visits, smoking status, BMI, and type-2 diabetes with likelihood of CRC screening did not differ significantly by HIV status (Table 2). A greater number of outpatient visits in the year before start of follow-up was associated with higher likelihood of CRC screening for both PWH and persons without HIV. Black race and current smoking were associated with lower likelihood of CRC screening for both PWH and persons without HIV (Table 2). We observed no difference in the type of first screening test received by PWH and persons without HIV (see Table 1, Supplemental Digital Content, http://links.lww.com/QAI/B300). FIT was the most common first CRC screening test for both groups (83.0% in PWH and 83.6% in persons without HIV).

FIGURE 1.
FIGURE 1.:
Time to first CRC screening and 95% confidence bands, by HIV status—Kaiser Permanente Northern California, 2005–2016.
TABLE 2.
TABLE 2.:
Factors Associated With CRC Screening, by HIV Status—Kaiser Permanente Northern California, 2005-2016

Screening Outcomes

Among 1844 PWH and 16,098 persons without HIV whose first screening event was a FIT, 5.9% and 5.1% had a positive result (P = 0.17). Among 822 PWH and 6619 persons without HIV with a sigmoidoscopy or colonoscopy, adenoma was detected in 161 (19.6%) PWH and 1498 (22.6%) persons without HIV (P = 0.052), and CRC was detected in 4 (0.5%) PWH and 69 (1.0%) persons without HIV (P = 0.14; Table 3). Of the cancers detected, early stage (I or II) cancer was detected in 3 (75%) PWH and 51 (74%) persons without HIV (P = 0.21); advanced stage (III or IV) cancer was detected in 1 (25%) PWH and 18 (26%) persons without HIV (P = 0.43). CRC was detected in the proximal colon for 1 (25%) PWH and 14 (20%) persons without HIV, in the distal colon for 2 (50%) PWH and 31 (45%) persons without HIV, and in the rectum for 1 (25%) PWH and 24 (35%) persons without HIV. None of the rectal CRC tumors were of squamous cell histology. All CRC in PWH and most (49 out of 69) CRC in persons without HIV were detected via colonoscopy.

TABLE 3.
TABLE 3.:
Prevalence of Adenoma* and Invasive Colorectal Cancer† at Sigmoidoscopy or Colonoscopy CRC Screening, by HIV Status—Kaiser Permanente Northern California, 2005–2016

We found suggestion of a lower prevalence of adenoma and CRC among PWH, which only reached statistical significance in unadjusted models (unadjusted PR: 0.86, 95% CI: 0.75 to 1.00, P = 0.049; adjusted PR: 0.97, 95% CI: 0.83 to 1.12, P = 0.13). We also found suggestion of higher prevalence of adenoma and CRC among some subgroups of PWH, including patients who were underweight or who had IBD; however, these findings did not reach statistical significance (Table 4). The associations of sex, age, race/ethnicity, number of outpatient visits, smoking, and type-2 diabetes with prevalence of adenoma and CRC were not different by HIV status (Table 4). Among PWH, we observed no difference in adenoma or CRC detection by CD4 count, duration of HIV infection, HIV RNA level, use of antiretroviral medications, or previous AIDS diagnosis (see Table 2, Supplemental Digital Content, http://links.lww.com/QAI/B300).

TABLE 4.
TABLE 4.:
Factors Associated With Detection of Adenoma or Invasive Colorectal Cancer at Sigmoidoscopy or Colonoscopy CRC Screening, by HIV Status—Kaiser Permanente Northern California, 2005–2016

Sensitivity Analyses

Our sensitivity analyses of patients for whom follow-up started at age 50 included 1630 PWH and 14,742 persons without HIV. No differences in results were observed in this subset of the full cohort for any of our analyses (results not shown), so we present the results of our primary analyses for only the full cohort.

DISCUSSION

In this study of CRC screening rates and outcomes in PWH and persons without HIV, we found that in a setting with an organized CRC screening program, overall screening uptake was high, with 86% of PWH and 79% of persons without HIV completing CRC screening within 5 years of follow-up. PWH were more likely to be screened and to be screened sooner, and there was no significant difference in clinical and demographic predictors of screening or type of first CRC screening test among PWH compared with persons without HIV. Of particular importance, we found that there was no significant difference in adenoma and CRC prevalence, or stage at CRC diagnosis in PWH (who were at otherwise average risk for CRC) compared with their uninfected counterparts at first CRC screening. In fact, we observed a slightly lower prevalence of adenoma and CRC among PWH, although these findings were not significant in adjusted analyses and the total number of cancers detected were small (n = 4). These results suggest that the age of CRC screening initiation recommended for the general population is also suitable for the HIV population. Furthermore, we found little evidence that patients with more severe HIV disease would require earlier screening because HIV-specific clinical factors associated with HIV disease severity (eg, CD4 level, nadir CD4 count, HIV RNA level, and history of AIDS) did not influence likelihood of adenoma or CRC detection. Also, known risk factors for CRC (eg, older age, Black race, smoking, and type-2 diabetes) had similar associations with adenoma and CRC detection for both PWH and persons without HIV.

Our finding that PWH were more likely to obtain CRC screening than persons without HIV differed from other studies,13,18,21,28–31 which reported under-use of CRC screening in HIV populations compared with the general population, even among patients in HIV care. The high CRC screening rate observed in our study population, which exceeds both national and state screening rates reported for the general population,32 can be largely attributed to KPNC's combined opportunistic and organized screening approach.4 To ensure high compliance with CRC screening recommendations within KPNC, screening-eligible patients receive mail, email, and telephone reminders, and those who are not up-to-date with screening are mailed FIT kits annually.4 In addition, electronic health record-based prompts remind health care providers, including HIV and infectious disease specialists, to offer screening at all clinical encounters for screening-eligible patients.4 The steep increase in cumulative proportion of patients screened within 1 year of follow-up reflects KPNC's adherence to the Healthcare Effectiveness Data and Information Set health care quality metric, which aims to provide CRC screening to all eligible patients by the end of the year of their 51st birthday. This explains the high proportion of subjects who initiated CRC screening during the recommended age range, with particularly high uptake in the first year of follow-up.

Some previous studies have linked screening uptake among PWH to the frequency of their care usage as patients who have contact with a primary care provider may be more likely to be offered or to accept screening.21,33–35 In this study, the greater likelihood of CRC screening among PWH compared with persons without HIV was not explained by a higher number of outpatient visits, which was associated with higher CRC screening uptake for all patients regardless of HIV status. Although provider–patient discussions about CRC screening may indeed increase screening uptake35,36 and endoscopic follow-up after a positive FIT result,33 our results show that proactive, targeted outreach via mailed FIT kits can be a key strategy for achieving broad initial CRC screening uptake in screening-naive populations without relying solely on in-office encounters. Similar to the findings in this study, a previous study of prostate cancer screening in our health care network reported higher screening rates among PWH compared with persons without HIV and attributed this to equal access to care,37 supporting the idea that systems-level factors that promote access to preventive health services can be essential for reducing disparities in cancer screening.

Consistent with other studies, we found that HIV infection was not associated with a significantly higher likelihood of CRC.6 In addition, adenoma prevalence was not significantly different in PWH and persons without HIV, suggesting that early detection and removal of precancerous adenomas may not explain why PWH have lower7,8 or comparable6 CRC incidence compared with the general population, but higher incidence of several other non-AIDS-defining cancers.11,38 These findings corroborate the conclusions of a recent study,7 which looked at CRC screening results by cancer stage and concluded that biological reasons and not screening effects may be a more likely explanation for lower than expected CRC among PWH.

In analyses restricted to PWH only, we found that HIV-specific clinical factors, including CD4 count (proxy for HIV disease severity), were not associated with increased CRC detection, which supports the notion that current guidelines to initiate CRC screening at age 50 apply equally to persons with and without HIV. Our study also demonstrates that high uptake of preventive CRC screening is possible in PWH and that disparities in uptake of CRC screening observed in some other settings were not present within an integrated setting with organized CRC screening outreach, which will be increasingly important as a greater number live beyond 50 years of age and the need for CRC screening increases. Further study will be needed to examine whether recommendations for the general population regarding frequency of screening, as well as the age at which to discontinue screening, both of which currently vary across published recommendations, would be applicable and appropriate for PWH. Previous research suggest that some providers screen PWH more frequently and more often for clinically indicated reasons, possibly because of their higher prevalence of gastrointestinal symptoms and anorectal disease.6,13,18,21 Therefore, research to determine optimal CRC frequency may be needed to minimize screening burden and maximize the impact of screening on CRC prevention among PWH.

This study had several limitations. First, in building our study cohort of persons without previous CRC screening, it was not possible to comprehensively ascertain CRC screening history, particularly tests obtained at non-KPNC facilities or those that were obtained before commencement of KPNC membership. However, identification of eligible study members included a review of health records with verification of CRC screening history and CRC diagnosis in a linked cancer registry. Additionally, sensitivity analyses examining those who started follow-up at age 50 yielded results consistent with those of the full cohort, suggesting that our results were not influenced by patients who may have had previous screenings not captured in the KPNC health records. Second, we cannot exclude the possibility of missed or delayed adenoma or CRC diagnoses in subjects who had a positive FIT result, but did not return for endoscopic screening during the study period. However, the percentage of FIT+ tests were comparable in patients with and without HIV, so this limitation is unlikely to affect our comparisons of adenoma and CRC findings across HIV groups. Third, other CRC screening tests including CT colonography, double-contrast barium enema, and fecal occult blood tests are not routinely ordered at KPNC, and were not included in the study. Finally, although we accounted for key demographic and clinical CRC risk factors, we were unable to account for some behavioral factors such as aspirin use, diet, and exercise, because they were not captured in electronic records-based data.

Our study design had several advantages over previous studies, including exclusion of high-risk CRC patients and CRC screens conducted for diagnostic purposes, as well as inclusion of demographically matched comparator persons without HIV. Our analyses benefited from the high overall uptake of CRC screening in our study population, which enabled evaluation of a large number of screening outcomes and generated more reliable estimates of adenoma and CRC prevalence than previous studies with low or disproportionate CRC screening uptake across groups. In addition, the type of screening tests used was similar by HIV status, which enabled more accurate comparison of adenoma and CRC detection.

CONCLUSIONS

In an integrated health care system with overall high CRC screening uptake due to an organized mailed FIT-based outreach program, time to first screen, factors associated with screening, screening modality, and outcomes of screening were similar in PWH and persons without HIV. HIV-specific factors were not associated with increased likelihood of adenoma or CRC, and traditional risk factors for CRC did not seem to have a differential effect on adenoma or CRC detection by HIV status. Therefore, we found no evidence that PWH need different CRC screening guidelines from the general population.

REFERENCES

1. Shiels MS, Pfeiffer RM, Gail MH, et al. Cancer burden in the HIV-infected population in the United States. J Natl Cancer Inst. 2011;103:753–762.
2. Cronin KA, Lake AJ, Scott S, et al. Annual report to the nation on the status of cancer. Part I: National cancer statistics. Cancer. 2018;124:2785–2800.
3. Bevan R, Rutter MD. Colorectal cancer screening-who, how, and when? Clin Endosc. 2018;51:37–49.
4. Levin TR, Corley DA, Jensen CD, et al. Effects of organized colorectal cancer screening on cancer incidence and mortality in a large community-based population. Gastroenterology. 2018;155:1383–1391.e5.
5. Bénard F, Barkun AN, Martel M, et al. Systematic review of colorectal cancer screening guidelines for average-risk adults: summarizing the current global recommendations. World J Gastroenterol. 2018;24:124–138.
6. O'Neill TJ, Nguemo JD, Tynan AM, et al. Risk of colorectal cancer and associated mortality in HIV: a systematic review and meta-analysis. J Acquir Immune Defic Syndr. 2017;75:439–447.
7. Coghill AE, Engels EA, Schymura MJ, et al. Risk of breast, prostate, and colorectal cancer diagnoses among HIV-infected individuals in the United States. J Natl Cancer Inst. 2018;110:959–966.
8. Hernandez-Ramirez RU, Shiels MS, Dubrow R, et al. Cancer risk in HIV-infected people in the USA from 1996 to 2012: a population-based, registry-linkage study. Lancet HIV. 2017;4:e495–e504.
9. Coghill AE, Shiels MS, Rycroft RK, et al. Rectal squamous cell carcinoma in immunosuppressed populations: is this a distinct entity from anal cancer? AIDS. 2016;30:105–112.
10. Marcus JL, Chao C, Leyden WA, et al. Survival among HIV-infected and HIV-uninfected individuals with common non-AIDS-defining cancers. Cancer Epidemiol Biomarkers Prev. 2015;24:1167–1173.
11. Silverberg MJ, Chao C, Leyden WA, et al. HIV infection, immunodeficiency, viral replication, and the risk of cancer. Cancer Epidemiol Biomarkers Prev. 2011;20:2551–2559.
12. Sigel C, Cavalcanti MS, Daniel T, et al. Clinicopathologic features of colorectal carcinoma in HIV-positive patients. Cancer Epidemiol Biomarkers Prev. 2016;25:1098–1104.
13. Guest JL, Rentsch CT, Rimland D. Comparison of colorectal cancer screening and diagnoses in HIV-positive and HIV-negative veterans. AIDS Care. 2014;26:1490–1493.
14. Ford RM, McMahon MM, Wehbi MA. HIV/AIDS and colorectal cancer: a review in the era of antiretrovirals. Gastroenterol Hepatol. 2008;4:274–278.
15. Wasserberg N, Nunoo-Mensah JW, Gonzalez-Ruiz C, et al. Colorectal cancer in HIV-infected patients: a case control study. Int J Colorectal Dis. 2007;22:1217–1221.
16. Bini EJ, Green B, Poles MA. Screening colonoscopy for the detection of neoplastic lesions in asymptomatic HIV-infected subjects. Gut. 2009;58:1129–1134.
17. Nayudu SK, Balar B. Colorectal cancer screening in human immunodeficiency virus population: are they at average risk? World J Gastrointest Oncol. 2012;4:259–264.
18. Reinhold JP, Moon M, Tenner CT, et al. Colorectal cancer screening in HIV-infected patients 50 years of age and older: missed opportunities for prevention. Am J Gastroenterol. 2005;100:1805–1812.
19. Berretta M, Cappellani A, Di Benedetto F, et al. Clinical presentation and outcome of colorectal cancer in HIV-positive patients: a clinical case-control study. Onkologie. 2009;32:319–324.
20. Kasapovic A, Boesecke C, Schwarze-Zander C, et al. Screening colonoscopy in HIV-infected patients: high rates of mucosal abnormalities in a German HIV-infected cohort. HIV Med. 2014;15:175–181.
21. Iqbal S, Browne-McDonald V, Cerulli MA. Recent trends for colorectal cancer screening in HIV-infected patients. Dig Dis Sci. 2010;55:761–766.
22. Bini EJ, Park J, Francois F. Use of flexible sigmoidoscopy to screen for colorectal cancer in HIV-infected patients 50 years of age and older. Arch Intern Med. 2006;166:1626–1631.
23. Gordon NP. Similarity of the Adult Kaiser Permanente Membership in Northern California to the Insured and General Population in Northern California: Statistics from the 2011 California Health Interview Survey: Kaiser Permanente Northern California Division of Research; 2015. Available at: https://divisionofresearch.kaiserpermanente.org/projects/memberhealthsurvey/SiteCollectionDocuments/chis_non_kp_2011.pdf.
24. Fedewa SA, Corley DA, Jensen CD, et al. Colorectal cancer screening initiation after age 50 years in an organized program. Am J Prev Med. 2017;53:335–344.
25. Corley DA, Jensen CD, Marks AR, et al. Adenoma detection rate and risk of colorectal cancer and death. N Engl J Med. 2014;370:1298–1306.
26. Gail MH. Frequency matching. In: Encyclopedia of Biostatistics: John Wiley & Sons, Ltd; 2005.
27. Zou G. A modified Poisson regression approach to prospective studies with binary data. Am J Epidemiol. 2004;159:702–706.
28. Keller SC, Momplaisir F, Lo Re V, et al. Colorectal cancer incidence and screening in US medicaid patients with and without HIV infection. AIDS Care. 2014;26:716–722.
29. Momplaisir F, Mounzer K, Long JA. Preventive cancer screening practices in HIV-positive patients. AIDS Care. 2014;26:87–94.
30. Antoniou T, Jembere N, Saskin R, et al. A population-based study of the extent of colorectal cancer screening in men with HIV. BMC Health Serv Res. 2015;15:51.
31. Burkholder GA, Tamhane AR, Appell LE, et al. Short communication: viral suppression is associated with increased likelihood of colorectal cancer screening among persons living with HIV/AIDS. AIDS Res Hum Retroviruses. 2015;31:519–524.
32. Joseph DA, King JB, Richards TB, et al. Use of colorectal cancer screening tests by state. Prev Chronic Dis. 2018;15:E80.
33. Halm EA, Beaber EF, McLerran D, et al. Association between primary care visits and colorectal cancer screening outcomes in the era of population health outreach. J Gen Intern Med. 2016;31:1190–1197.
34. Campbell J, Young B. Use of screening colonoscopy in ambulatory HIV-infected patients. J Int Assoc Physicians AIDS Care (Chic). 2008;7:286–288.
35. Momplaisir F, Long JA, Badolato G, et al. The role of primary care physicians in improving colorectal cancer screening in patients with HIV. J Gen Intern Med. 2012;27:940–944.
36. Doubeni CA, Fedewa SA, Levin TR, et al. Modifiable failures in the colorectal cancer screening process and their association with risk of death. Gastroenterology. 2019;156:63–74.
37. Marcus JL, Chao CR, Leyden WA, et al. Prostate cancer incidence and prostate-specific antigen testing among HIV-positive and HIV-negative men. J Acquir Immune Defic Syndr. 2014;66:495–502.
38. Shiels MS, Cole SR, Kirk GD, et al. A meta-analysis of the incidence of non-AIDS cancers in HIV-infected individuals. J Acquir Immune Defic Syndr. 2009;52:611–622.
Keywords:

colorectal cancer; screening; colonoscopy; adenoma

Supplemental Digital Content

Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.