Community Health Centers’ Performance in Cancer Screening and Prevention

Abstract

Introduction:

Little is known about what clinic-level factors differentiate community health centers (CHCs) who achieve high performance on cancer preventive care metrics. This study aims to describe the longitudinal trends in the delivery of 3 cancer preventive care metrics (cervical and colorectal cancer screenings and tobacco-cessation intervention) and define and compare CHCs with high versus low cancer preventive care performance.

Methods:

This observational study used 2012–2019 CHC data (n=933) from the Uniform Data System. High/low performance was based on Healthy People 2020 targets and sample distribution. For each cancer preventive care metric, the percentage of CHCs that met high (≥70.5% at cervical or colorectal cancer screening or >80% tobacco-cessation intervention) and low thresholds at 1, 2, and all 3 screenings was estimated. Multivariable generalized estimating equations logistic regression modeling was used to assess the CHC-level factors associated with screening performance.

Results:

The CHCs’ performance for tobacco-cessation intervention remained steady at ≥80%, with a small increase over time. Performance for cervical cancer screening remained unchanged with about 50% of patients screened. Colorectal cancer screening performance increased over time from about 30% in 2012 to 44% in 2019. A very small number of CHCs reached high performance (3%) in all 3 indicators, and 13% of CHCs were high in any 2 of the outcomes in 2019. Higher patient volume, greater proportion of Hispanic patients, fewer uninsured patients, and CHCs located in the Northeast region were associated with high performance in 2019.

Conclusions:

Very few CHCs meet all Healthy People 2020 goals in cancer screenings and may struggle to achieve the 2030 goals. Very few indicators differentiated high and low performers.

INTRODUCTION

Community health centers (CHCs) provide services to >29 million people, predominantly low-income, racial and ethnic minority patients across the U.S. each year regardless of their ability to pay.¹ CHCs offer high-quality health care, reduce barriers to cost, and accept patients without insurance.^1–3 However, although CHCs exceed national standards for the delivery of many healthcare services (e.g., diabetes and hypertension care), cancer preventive care services vary across CHCs, with some meeting Healthy People 2020 targets, and others not.^4–8

Many challenges face CHCs in delivering cancer preventive care. CHCs are often underfunded, short staffed,⁹ prone to staff turnover, and lack other resources found in integrated care delivery systems.^10,11 Additionally, CHCs serve a high-need patient population with complex problems such as health insurance instability, exposure to adverse social determinants of health (e.g., food/housing insecurity), and are often in poorer health (e.g., higher prevalence of multi-morbidity) than those seen outside CHCs,¹² impacting performance of cancer preventive care.¹¹ To date, little is known about what clinic-level factors differentiate CHCs who achieve high performance at 1, some, or all of these cancer preventive care services. Understanding performance patterns in CHCs can inform the development of targeted interventions aimed at improving adoption and spread of evidence-based practices. Therefore, the objectives of this study are 2-fold. The first is to describe longitudinal trends (2012–2019) in the delivery of cervical and colorectal cancer screenings, and tobacco-cessation interventions in a sample of nationally representative CHCs. It is hypothesized that performance in these metrics have increased over time but still fall short of Healthy People 2020 targets. Second, using 2019 data to capture the current state of the delivery of cancer preventive care (i.e., the year prior to the Healthy People 2020 end date), the aim is to compare CHCs that had high cancer preventive care performance to those with lower performance to identify CHC-level factors associated with performance. It is hypothesized that CHCs that perform well across the 3 cancer screening outcomes will have different characteristics than CHCs with poorer performance.

METHODS

Study Sample

This observational study used data from the Uniform Data System (UDS) for the time period of 2012–2019. The UDS is an annual reporting system a standardized set of information from CHCs (including federally qualified health centers) awarded funding by the Health Resources and Services Administration.⁵ The goal of the UDS reporting system is to improve CHC performance and operations and assess trends over time in performance measures. The UDS data include CHC-level standardized information on patient-level characteristics and healthcare utilization rates. Although UDS collects data from program lookalikes, there are no data for these centers in this study. The final analytic sample included 933 (of 1,383) CHCs with data across all years from 2012 to 2019 from 50 U.S. states and the District of Columbia.

Measures

Included were UDS data from 3 available cancer screening and prevention quality metrics: colorectal and cervical cancer screening, and tobacco-cessation intervention. These metrics follow the U.S. Preventive Services Task Force recommendations.¹³ Appendix Table 1 provides details. Briefly:

1. Colorectal cancer screening is the percentage of patients aged 50–75 years with a medical visit in the reporting year who had appropriate screenings for colorectal cancer (e.g., fecal occult blood test, fecal immunochemical test, colonoscopy).

2. Cervical cancer screening is the percentage of women aged 21–64 years who had cervical cytology performed every 3 years, or aged 30–64 years who had cervical cytology/human papillomavirus co-testing performed every 5 years.

3. Tobacco-cessation intervention is the percentage of patients aged ≥18 years seen for ≥2 medical visits in the reporting year or ≥1 preventive medical visit during the measurement period who were screened for tobacco use ≥1 time within a 24-month period and received a tobacco-cessation intervention (either counseling or pharmacotherapy) if identified as a tobacco user.

For colorectal and cervical cancer screenings, performance was available for the years 2012–2019. For tobacco-cessation intervention, data were available from 2014 to 2019 as reporting requirements prior to 2014 focused on screening only rather than receiving a tobacco-cessation intervention.

Definitions for high and low performance for each cancer metric were based on Healthy People 2020 targets and on the observed distribution of performance in the sample. First, high performance in colorectal and cervical cancer screenings was defined as those with a quality metric of ≥70.5% of screened eligible patients (the Healthy People 2020 goal for colorectal cancer screening¹⁴). Though Healthy People 2020 goal for cervical cancer screening is 93%, the colorectal cancer screening threshold was used because only 1 of the 933 CHCs (0.1%) had met the cervical cancer screening goal. As noted previously, CHCs serve a complex patient population and such a goal would limit the information gained among those who reach a relatively high performance (≥70.5%). High-performing CHCs in tobacco-cessation intervention were those meeting the Healthy People 2020 goal of ≥80%. Lastly, CHCs were assigned the label of “high performer” in overall cancer preventive care if they reached these thresholds at a minimum of 2 of the 3 cancer metrics.

Low performers were defined as those with a quality metric <25th percentile of the sample distribution for colorectal (<32.4% of screened patients) and cervical (<41.6% of screened patients) cancer screenings, and 22nd percentile for tobacco (owing to the high percentage of CHCs reaching rates ≥80%). The threshold for the tobacco metric was <80%. Lastly, CHCs were assigned the label of “low performer” in overall cancer preventive care if the CHC reached these lower thresholds on a minimum of 2 metrics.

The following independent CHC-level factors were available in the UDS data: total number of patients, percentage of patients who are adults (aged ≥18 years), percentage of uninsured patients, percentage on Medicaid/Children’s Health Insurance Program, percentage at ≤100% the federal poverty level, percentage of minority patients (includes all non-White race and Hispanic ethnicity), percentage of Hispanic patients, percentage of Black/African American patients, and U.S. region (South, Northeast, Midwest, and West¹⁵) where the CHC is located. All independent variables, except for region, were categorized into 4 groups using Jenks natural breaks optimization,^16–19 an approach that classifies the data into different groups according to the breaks that naturally exist in the data, to facilitate analysis and interpretation (e.g., examine potential nonlinear, or U-shaped relationships with cancer preventive care performance).

Statistical Analysis

Descriptive statistics were performed to evaluate longitudinal trends of cancer metrics. CHC characteristics were described in the total sample for the year 2019 to reflect current patient- and clinic-level characteristics of CHCs nationwide. Descriptive profile plots were used to describe and visualize changes in performance, longitudinal trends in the delivery of cervical and colorectal cancer screenings, and tobacco-cessation intervention performance for the full sample from 2012 to 2019 (2014 to 2019 for tobacco).

Next, 2019 data were used to describe the percentage of CHCs that were high or low on ≥2 metrics, and for all 3 metrics. A multivariable generalized estimating equations logistic regression model was used to compare CHCs that had high cancer preventive care performance with those with lower performance in 2019. The outcome was the binary indicator of overall high versus low performance on ≥2 metrics. Independent variables included those described previously. An exchangeable working correlation structure (i.e., assumes correlations between multiple observations are constant) was assumed to account for clustering of CHCs within states. Along with reporting ORs and 95% CIs, p-values were reported for trends to detect a monotonic increasing/decreasing trend within each independent variable except for U.S. region, where p-values were reported for comparing each region with the South region. Multicollinearity between the independent variables was not an issue in the modeling (variance inflation factor <10 for all variables). Analyses were performed using RStudio, version 1.3.959 and statistical significance was set at p<0.05. The study did not involve any individual-level data and the data are publicly available; thus, this study was exempted from formal IRB approval.

RESULTS

Table 1 describes the characteristics of 933 study CHCs for the year 2019. Most were located in either the South (36%) or the West (28%) U.S. region. The average number of patients seen per center was 23,743 patients with 73% of those patients being adults. More than half of patients served by these CHCs were racial/ethnic minorities (54%), >60% were ≤100% of the federal poverty level, and a quarter were uninsured (25%). Comparisons of CHCs that met study criteria with those that did not are reported in Appendix Table 2.

Table 1.

Community Health Center Characteristics in the U.S., 2019

Community health center characteristic	Mean (SD)
Number of health centers	933
Total patients	23,743 (27,411)
% of children (<18 years old)	27.1 (11.6)
% of adult patients	73.0 (11.6)
% racial and/or ethnic minority	54.4 (31.2)
% Hispanic/Latino ethnicity	27.4 (26.9)
% Black/African American	21.4 (25.4)
% Uninsured	24.5 (17.0)
% with Medicaid/CHIP	42.1 (17.6)
% with Medicare	11.7 (7.5)
% with other third-party (private) insurance	21.7 (12.6)
% of patients ≤200% of poverty	89.0 (11.7)
% of patients ≤100% of poverty	63.6 (18.3)
% of cervical cancer screenings	52.3 (16.1)
% of colorectal cancer screenings	43.8 (16.3)
% of tobacco cessation intervention	85.6 (12.0)
U.S. region, N (%)
South	335 (35.8)
Midwest	170 (18.2)
Northeast	170 (18.2)
West	260 (27.8)

CHIP, Children’s Health Insurance Program.

Figure 1 shows the trends in cancer preventive care metrics across the 933 CHCs by year. The performance for tobacco-cessation intervention remained >80% of screened eligible patients and slightly increased over time. In 2019, the average performance in tobacco-cessation intervention across CHCs was 86%. For cervical cancer screening, the average performance over time remained mostly unchanged with about 50% of eligible patients screened across CHCs (2019 average=52%). The percentage of eligible adults screened for colorectal cancer increased over time from about 30% in 2012 to 44% in 2019.

Figure 1.

Trends in the performance of cancer preventive care metrics across 933 U.S. community health centers, 2012‒2019.

Notes: Colorectal cancer screening is the percentage of patients aged 50‒75 years with a medical visit in reporting year who had appropriate screenings for colorectal cancer (e.g., FOBT, FIT, colonoscopy) following the USPSTF recommendation. Cervical cancer screening is the percentage of women aged 21‒64 years who had cervical cytology performed every 3 years, or aged 30‒64 years who had cervical cytology/human papillomavirus (HPV) co-testing performed every 5 years. Tobacco cessation intervention is the percentage of patients aged 18 years and older who were screened for tobacco use one or more times within a 24-month period and who received tobacco cessation intervention if identified as a tobacco user.

FOBT, fecal occult blood test; FIT, fecal immunochemical test; USPSTF, U.S. Preventive Services Task Force.

As shown in Table 2, most CHCs were classified as high performers for tobacco-cessation intervention (78%); however, only 12% of CHCs were high in cervical cancer screening performance, and even fewer were high in colorectal cancer screening performance (5%). When evaluating overall cancer preventive care, 13% of clinics were high in any 2 of the metrics and only 3% were high in all 3.

Table 2.

Number of High and Low Performing Health Systems From 933 Community Health Centers in the U.S., 2019

Cancer preventive metrics	High performersa	Low performersb
Cervical cancer screening, N (%)	112 (12)	234 (25)
Colorectal cancer screening, N (%)	47 (5)	234 (25)
Tobacco cessation intervention, N (%)	726 (78)	207 (22)
In any 2 cancer preventive care metrics, N (%)	122 (13)	188 (20)
In all 3 cancer preventive care metrics, N (%)	23 (3)	59 (6)

^aTo be considered high performing, health centers must have had performance of ≥70.5% for cervical and colorectal cancers screening rates and ≥80% for tobacco cessation intervention rates.

^bLow performing centers are determined if their quality metric was below the 25th percentile of the sample distribution for colorectal and cervical cancer screening rates and bellow the 22nd percentile for tobacco cessation intervention rates. This equated to <32.4% for colorectal cancer screening, <41.6% for cervical cancer screening and <80% for tobacco cessation intervention.

The descriptive characteristics of high- and lower-performing CHCs for at least 2 of the metrics in 2019 and ORs derived from the generalized estimating equations logistic regression model are shown in Table 3. Overall, CHCs with a higher number of patients had higher odds of being a high performer (p<0.001 for trend). CHCs with a greater proportion of Hispanic patients (>68%) were 4.51 times more likely to be higher performers (OR=4.51, 95% CI=1.45, 13.97). Also, Northeast region CHCs had better performance than the South region (OR=3.30, 95% CI=1.19, 9.15). CHCs with a large proportion of uninsured patients were significantly less likely to be high performers (OR=0.12, 95% CI=0.02, 0.94). Other patient panel characteristics were not associated with performance.

Table 3.

Community Health Center Factors Associated With Being a High Performer in Overall Cancer Preventive Care, 2019

CHC characteristics	High performersa N (%)	Low performersb N (%)	OR (95%CI)	P for Trend
N of CHCs	122	188
Total patients
<13,331	41 (33.6)	127 (67.6)	ref	<0.001
13,331–59,668	45 (36.9)	45 (23.9)	3.15 (1.44, 6.92)
59,668–113,563	24 (19.7)	13 (6.9)	6.31 (3.04, 13.10)
>113,563	12 (9.8)	3 (1.6)	7.32 (1.93, 27.77)
Adult (aged ≥18 years)
<61.3%	28 (23.0)	21 (11.2)	ref	0.663
61.3%−74%	49 (40.2)	60 (31.9)	0.61 (0.18, 2.03)
74.1%−84.9%	31 (25.4)	62 (33.0)	0.75 (0.20, 2.72)
85%−100%	14 (11.5)	45 (23.9)	0.63 (0.16, 2.44)
Racial and ethnic minority
<31%	27 (22.1)	55 (29.3)	ref	0.211
31%−59.5%	11 (9.0)	43 (22.9)	0.67 (0.19, 2.42)
59.6%−81.8%	34 (27.9)	48 (25.5)	2.36 (0.70, 7.95)
81.8%−100%	50 (41.0)	42 (22.3)	1.91 (0.59, 6.16)
Hispanic/Latino Ethnicity
<16.6%	46 (37.7)	108 (57.4)	ref	0.035
16.6%−39.6%	15 (12.3)	46 (24.5)	0.84 (0.28, 2.48)
39.7%−68.1%	28 (23.0)	21 (11.2)	1.47 (0.50, 4.29)
68.2%−100%	33 (27.0)	13 (6.9)	4.51 (1.45, 13.97)
Black/African American
<12.1%	65 (53.3)	94 (50.0)	ref	0.359
12.1%−33.5%	30 (24.6)	32 (17.0)	1.36 (0.28, 6.69)
33.6%−60.6%	14 (11.5)	32 (17.0)	0.48 (0.09, 2.53)
60.7%−100%	13 (10.7)	30 (16.0)	0.44 (0.08, 2.48)
Uninsured
<18.8%	54 (44.3)	70 (37.2)	ref	0.253
18.8%−36.4%	41 (33.6)	75 (39.9)	0.6 (0.31, 1.15)
36.5%−62.2%	25 (20.5)	25 (13.3)	1.1 (0.45, 2.66)
62.3%−100%	2 (1.6)	18 (9.6)	0.12 (0.02, 0.94)
Medicaid/CHIP
<24.8%	17 (13.9)	57 (30.3)	ref	0.894
24.8%−41.7%	35 (28.7)	54 (28.7)	1.38 (0.60, 3.17)
41.8%−58.4%	38 (31.1)	38 (20.2)	1.52 (0.40, 5.80)
58.5%−100%	32 (26.2)	39 (20.7)	0.81 (0.12, 5.38)
Patients ≤100% of FPL
<43%	16 (13.1)	26 (13.8)	ref	0.228
43%−64.3%	38 (31.1)	56 (29.8)	0.68 (0.27, 1.72)
64.4%−81%	49 (40.2)	64 (34.0)	0.55 (0.20, 1.45)
81.1%−100%	19 (15.6)	42 (22.3)	0.49 (0.15, 1.61)
U.S. regionc
South	37 (30.3)	83 (44.1)	ref
Midwest	19 (15.6)	29 (15.4)	1.09 (0.42, 2.80)	0.865
Northeast	37 (30.3)	24 (12.8)	3.30 (1.19, 9.15)	0.022
West	29 (0.24)	52 (0.28)	0.42 (0.11, 1.56)	0.194

Notes: Bolded estimates were statistically significant at p<0.05. For this binary outcome of high vs low in at least 2 outcomes, authors estimated ORs of being a ‘high performing’ CHC using multivariable generalized estimating equations (GEE) logistic regression modeling assuming an exchangeable working correlation structure to account for clustering of CHCs within states. Table reports p-values for trends to detect a monotonic increasing/decreasing trends within each independent variable.

^aTo be considered high performing in overall cancer preventive care, health centers must have had high performing thresholds at a minimum of 2 of the 3 cancer preventive care metrics (≥70.5% for cervical and colorectal cancer screenings and ≥80% for tobacco cessation intervention).

^bLow performing centers are determined if their quality metric was below the 25th percentile of the sample distribution for colorectal and cervical cancer screenings and bellow the 22nd percentile for tobacco cessation intervention. This equated to <32.4% for colorectal cancer screening, <41.6% for cervical cancer screening and <80% for tobacco cessation intervention.

^cThe reported p-value is not a trend test as this is not an ordinal variable. The reported p-value is comparing each region to the south (reference group).

FPL, federal poverty level; CHC, community health center; CHIP, Children’s Health Insurance Program.

DISCUSSION

This study described longitudinal cancer preventive care from 2012 to 2019 and utilized the most recent available data (2019) to evaluate the current state of the delivery of cancer preventive care and how it meets Healthy People 2020 targets in U.S. CHCs. It should be emphasized that too few CHCs (n=1) met the Healthy People 2020 targets for cervical cancer screenings and as a result the colorectal cancer screening threshold of 70.5% was used to define high cervical cancer screening performance. Overall, a very small number of CHCs reached high performance on all 3 metrics. Similarly, very few CHCs had low performance on all 3 metrics, suggesting relatively little variation in performance across the vast majority of CHCs. Few of the measured characteristics differentiated those with high or low performance. CHCs in the Northeast and those with a larger proportion of Hispanic patients had higher performance than their counterparts. Previous studies have shown that Hispanic patients have higher rates of cancer screenings than non-Hispanic Whites,^6,20,21 which may explain the variation in CHCs’ performance. CHCs with low performance had fewer patients, but had a higher number of uninsured patients. Previous evidence suggests that patients without health insurance and with high social need are prioritizing non-healthcare needs (e.g., food, lodging) over preventive care,^22–24 which may impact CHC performance. Screening for social needs and connecting patients to community assistance is recognized as essential to improve overall health.²⁵ CHC patients in particular face many adverse social determinants that impede to their ability to seek needed preventive care, thus impacting clinical performance. Although CHCs screen for social needs more than non-CHC practices,²⁶ many barriers (e.g., inadequate reimbursement, resources)^27,28 to successful implementation and sustainability of social needs screening/referral exist. Health policies focusing on adequate reimbursement for screening and referral and financial assistance for staff training and resources are critical for sustaining this service.

A large proportion of CHCs had high performance for the tobacco-cessation intervention metric. CHCs serve populations disproportionately affected by cigarette smoking including low-income, Medicaid, and uninsured patients.²⁹ The increasing trend in provision of tobacco-cessation interventions in the early years of the study period could be related to the implementation of policies relevant to CHC patient populations. One was the implementation of the Centers for Medicare and Medicaid Services’ “meaningful use” of electronic health records in early to mid-2010s that provided incentive payments to eligible providers for improvements in care processes and outcomes, including metrics related to tobacco smoking assessment and provision of cessation interventions. Though the number of providers that participated in meaningful use from each CHC is unknown, most likely met one of the eligibility criteria: have ≥30% Medicaid patient volume, have ≥20% Medicaid patient volume for pediatrician, or practice primarily in a federally qualified or rural health center and have ≥30% patient volume attributable to needy individuals. Furthermore, electronic health records were modified to facilitate capture of smoking-related data. Previous studies show a positive association between this incentive program and smoking assessment and cessation assistance.^30,31 Additionally the Affordable Care Act may have impacted performance.^32–34 In January 2014, state Medicaid formularies were prohibited from excluding smoking-cessation medications and, among newly Affordable Care Act–eligible Medicaid enrollees, smoking-cessation treatments were required to be covered without cost sharing or prior authorization. Although not all states expanded Medicaid, the Affordable Care Act Medicaid expansions have made tobacco-cessation coverage available to millions of adult smokers who were not eligible for Medicaid or tobacco-cessation treatments pre-expansion.³⁵

Though rates of tobacco-cessation interventions were high overall, it is imperative that healthcare teams have the resources and systems that enable them to address cessation with all patients who use tobacco. Modifiable barriers to provide cessation assistance include lack of time, inadequate patient resources such as health insurance coverage, and inadequate skill in provision of smoking cessation.^35,36 Using a team-based approach (including referrals to outside resources) to reduce the time burden on clinicians, continued policy reform to ensure access to recommended cessation interventions regardless of insurance, and training in brief interventions could result in even higher rates of cessation assistance and increased quit attempts.

Colorectal cancer screening rates showed the largest improvement since 2012 but remain relatively low. This change may be related to specific grant-funded initiatives (e.g., Community Health Advocates Implementing Nationwide Grants for Empowerment and Equity⁸) that CHCs can utilize to implement strategies to improve colorectal cancer screening. It could also be associated with the increase in uptake of fecal immunochemical/fecal occult blood testing, shown to be effective in colorectal cancer screening.³⁷ A few interventions^38–43 (e.g., use of text messaging, use of practice facilitation, mailing of fecal immunochemical test, patient navigation program) have shown significant improvements in colorectal cancer screening rates and reducing patient barriers (e.g., transportation, scheduling assistance). Despite this improvement, colorectal cancer screening performance remains low. Barriers and challenges to colorectal cancer screening include but are not limited to: care coordination issues between testing facilities, lack of in-facility testing, patient knowledge and attitude toward screening, and barriers associated with implementing and sustaining effective interventions (hiring patient navigators, electronic health record modification, staff turnover).^39,41,42,44 Future initiatives need to identify strategies to deploy these successful interventions in all U.S. CHCs.

Cervical cancer screening is mainly completed within CHCs, eliminating some of the barriers (e.g., access to testing) that are faced with colorectal cancer screening; yet, no change in screening rates was observed over the study period. The revised Healthy People 2030 target for cervical cancer screening is lower than the 2020 target, rendering them potentially more achievable for CHCs; however, cervical cancer screening rates have remained the same in CHCs since 2012 and are significantly lower than the Healthy People 2030 goal (84.3%). In fact, previous findings⁴⁵ based on self-reported national data have noted declining trends in cervical cancer screenings, highlighting lack of health insurance as a major barrier to cervical cancer screening. Few targeted interventions in CHCs settings such as education intervention (targeting provider or patient knowledge^46,47) and patient reminders systems^47,48 have had some success in improving cervical cancer screening adherence but many barriers to cervical cancer screenings remain (e.g., clinician/patient ratio, burnout).^11,47,49 There is a need for innovative multilevel interventions to improve these rates.

Limitations

This study has some limitations. UDS reporting provides an average performance across CHC delivery sites. Many CHCs have multiple delivery sites that may have variable performance, patient characteristics, or be located in urban or rural areas. Future research is needed to determine whether the performance at the CHCs is similar across practices within the center. If not, practice-level changes should be evaluated to determine the variation between practices within CHCs and whether this has changed over time. Further, practice-level analyses could provide sufficient power to characterize performance for each cancer metric and other indicators not available in the UDS data. Additionally, although U.S. region was included in the analyses, within-region or state-level differences in performance could not be evaluated, which requires future research considerations. Moreover, this study could not differentiate between CHCs that are federally qualified health centers and those that are not. The tobacco-cessation intervention metric lacks details on how counseling is defined; thus, it is difficult to compare to non-UDS data and the UDS data do not differentiate screening rates from cessation intervention. Next, UDS data include a limited number of clinic-level indicators and other unmeasured factors (e.g., being part of a patient-centered medical home, funding, turnover) that may contribute to variation in screening rates. Lastly, how the quality measures were assessed in the reported of UDS data (e.g., chart review, derived from electronic health record) could not be ascertained, which could have an impact on measurement accuracy.

CONCLUSIONS

Despite these limitations, the findings highlight significant progress over time in CHCs’ colorectal cancer screening and tobacco-cessation intervention performance. The findings also showed that very few CHCs meet all Healthy People 2020 goals in cancer preventive care and may continue to struggle in achieving the 2030 goals without targeted interventions and enhanced resources. Lastly, very few clinic-level indicators differentiated high and low performers. These findings will be instrumental in providing a foundation for future research to understand performance in cancer preventive care of CHCs at the clinic level and facilitate the implementation of targeted interventions to improve cancer preventive care.