Adherence to Repeat Fecal Occult Blood Testing in an Urban Community Health Center Network
- First Online:
- Cite this article as:
- Liss, D.T., Petit-Homme, A., Feinglass, J. et al. J Community Health (2013) 38: 829. doi:10.1007/s10900-013-9685-x
- 233 Views
Annual fecal occult blood testing (FOBT) has the potential to reduce colorectal cancer mortality, but in practice it is challenging to complete FOBT every year. Repeat FOBT adherence may be especially low in community health center (CHC) settings, where many patients face barriers to annual FOBT completion. We conducted a retrospective cohort analysis to investigate adherence to annual FOBT in an urban CHC network that serves a predominantly Spanish-speaking, uninsured adult patient population. This study used data from the two-year period between January 2010 and December 2011, and included adults aged 50–74 who completed a screening FOBT with a negative result during the first 6 months of 2010. We examined whether each patient completed a second FOBT between 9 and 18 months after the initial negative FOBT, and tested whether repeat FOBT adherence was associated with patient characteristics or the number of clinic visits after the initial negative FOBT. Only 69 of 281 included patients completed repeat FOBT (24.6 % adherence), and none of 62 patients (0 %) with 0 clinic visits completed repeat FOBT. We detected no significant differences in adherence by age, sex, preferred language, insurance status, or number of chronic conditions. In multivariable regression, the adjusted relative risk of repeat FOBT was 1.66 (95 % CI 1.09–2.54; p = 0.02) among patients with 3 or more clinic visits (referent: patients with 1–2 visits). The observed low rate of adherence greatly diminishes the effectiveness of FOBT in reducing CRC mortality. Findings demonstrate the need for systems-based interventions that increase adherence without requiring face-to-face encounters.
KeywordsColorectal cancerCancer screeningAdherencePreventive careDisparities
Population-wide colorectal cancer (CRC) screening has the potential to substantially reduce CRC mortality [1, 2]. The U.S. Preventive Services Task Force (USPSTF) recommends screening for CRC via annual fecal occult blood testing (FOBT), sigmoidoscopy every 5 years (plus FOBT every 3 years), or colonoscopy every 10 years , citing approximately equal mortality reductions across approaches when assuming 100 % adherence [1, 2]. However, the true effectiveness of each screening modality depends on the rate of adherence to each screening regimen.
In practice, adherence to an annual FOBT screening regimen is challenging [4, 5]. Though less invasive than endoscopy, FOBT needs to be conducted annually [2, 6], or at least biennially , to lower mortality. There are few studies of longitudinal FOBT in the United States, but early evaluations detected suboptimal repeat screening rates among insured patients in integrated delivery systems. In Washington State, 44 % of eligible patients completed repeat FOBT , while a national study found that 14 % of veterans exclusively using FOBT completed at least four tests over 5 years .
Repeat FOBT adherence may be especially low in community health center (CHC) settings. CHCs often use FOBT for CRC screening, due to limited access and the high cost of endoscopy. However, many patients cared for by CHCs face barriers to completing FOBT every year, including lack of insurance , non-financial barriers to care (e.g., lack of transportation or child care) [9, 10], low health literacy  or low perceived need for screening . Although annual FOBT in vulnerable populations has the potential to play an important role in reducing persistent disparities in CRC mortality by race/ethnicity [13, 14] and socioeconomic status [15, 16], we are aware of no prior studies of adherence to annual FOBT in a CHC setting.
We conducted this study to investigate patients’ adherence to annual FOBT in a CHC network with a well-established program to increase CRC screening. Our objectives were to identify the proportion of eligible CHC patients who completed repeat FOBT within 18 months of an initial negative FOBT result, and to identify factors associated with repeat FOBT completion.
This study uses data from patients cared for at Erie Family Health Center (EFHC), a federally qualified health center (FQHC) network in the Chicago region serving a predominantly Spanish-speaking patient population; the majority of adult EFHC patients are poor and uninsured. From 2007 to 2009, EFHC implemented a multicomponent quality improvement initiative to increase CRC screening rates, including performance measurement, provider financial incentives, and standing orders for medical assistants to distribute guaiac card FOBTs during in-person visits. Endoscopy is not offered on-site at EFHC, but qualifying patients with a positive FOBT result have been able obtain no-cost colonoscopy at Northwestern Memorial Hospital (Chicago, IL) since 2009.
Study Design and Data Collection
We conducted a retrospective cohort analysis of patient-level data from the two-year period between January 2010 and December 2011. Information on patient demographics, health insurance, medical conditions, CRC screening and health care use was obtained through queries of EFHC’s Centricity electronic health record (General Electric). Study protocols were approved by Northwestern University’s institutional review board and EFHC’s research review committee.
Patients were included in this study if they were age 50–74 on January 1, 2010, had a preferred language of Spanish or English, and completed a screening FOBT with a negative result (e.g., three negative samples recorded) during the first 6 months of 2010. Patients were excluded if CRC screening was inappropriate, as indicated by an International classification of Diseases, Ninth Revision, clinical modification (ICD-9-CM) diagnosis or Current Procedural Terminology (CPT) code for total colectomy (ICD-9-CM 45.8; CPT 44150–44153, 44155–44156, 44210–44212). We also excluded individuals with any ICD-9-CM diagnosis for a condition where FOBT may be inappropriate, including: chronic diarrhea (787.91); Crohn’s disease (555.xx); ulcerative colitis (556.xx); iron deficiency anemia (280.9); gastrointestinal bleeding (578.xx); colonic polyp (211.3), or; malignant neoplasm of colon and other specified sites of colon and large intestine (153.X, 154.0, 154.1, 197.5, V10.05). In addition, we excluded 17 patients who underwent colonoscopy (CPT 45.23) within the past 10 years, as they had no need to conduct FOBT screening during this study.
The primary outcome was a binary measure of whether each patient completed a second FOBT between 9 and 18 months after the initial negative FOBT in 2010. FOBTs completed 9–12 months after initial negative tests were counted as instances of outcome achievement because we found that some providers distributed tests at visits preceding patients’ annual FOBT due dates.
We created measures describing individuals’ age on January 1, 2010 (50–54, 55–59, 60–64, 65–74), sex and preferred language (Spanish vs. English). Insurance coverage was dichotomized as yes/no. We created a summary score for the number of comorbid chronic illnesses (0, 1, ≥2) identified from a list of ten potential conditions: eight from the Dartmouth Atlas of Health Care (cancer, chronic pulmonary disease, coronary artery disease, congestive heart failure, peripheral vascular disease, severe chronic liver disease, renal failure, and dementia) , diabetes and hypertension. We determined the number of opportunities to offer repeat screening by measuring the number of visits to EFHC (0, 1, 2, 3, ≥4) between 9 and 18 months after the initial negative FOBT.
We calculated descriptive statistics for each patient characteristic and the percent of included patients who completed repeat FOBT. We conducted Pearson’s chi squared tests to examine group-level differences in repeat FOBT rates for measures of patient characteristics and EFHC clinic visits.
Finally, we estimated a multivariable Poisson regression model, incorporating robust variance estimates, to determine adjusted associations between repeat FOBT adherence and each independent variable (patient characteristics and the number of clinic visits) . After a review of adherence data found that patients with zero EFHC visits universally failed to complete repeat FOBT, these individuals were omitted from the regression sample (inclusion of these patients would have precluded model convergence). The regression model included a collapsed indicator of EFHC visits (1–2 vs. ≥3) due to the small number of patients in each remaining category of the original visit measure. Analyses were conducted using Stata, version 12 (College Station, TX).
Patient characteristics and repeat FOBT completion frequencies and rates, according to patient characteristics
Repeat FOBT completion, N (row %)
Only 69 patients completed repeat FOBT between 9 and 18 months after their initial negative FOBT (24.6 % adherence). Chi square tests detected no significant differences in repeat FOBT adherence by age, sex, preferred language, insurance status or number of chronic conditions (Table 1).
Multivariable Poisson regression included 219 patients who made any EFHC clinic visits, 31.5 % of whom completed repeat FOBT. Compared to patients with 1–2 clinic visits between 9 and 18 months after initial negative FOBT, patients with 3 or more visits had a 1.66 adjusted relative risk of repeat FOBT (95 % CI 1.09–2.54; p = 0.02). The regression model estimated no statistically significant associations between repeat FOBT and individual patient characteristics.
We found a very low rate of repeat FOBT adherence in this patient population. Only one-fourth of eligible patients completed repeat FOBT, despite the fact that the CHC network under study had already implemented a multicomponent initiative to increase CRC screening. This low rate of adherence greatly diminishes the effectiveness of FOBT in reducing CRC mortality.
The 24.6 % overall adherence to repeat FOBT screening in this CHC network (over 18 months) is substantially lower than the 44 % rate observed in a private integrated health plan and delivery system (over 24 months) . This difference may be attributable to factors not directly examined in this study, such as barriers to health care access [8, 19] or literacy level . In this study, however, completion of repeat FOBT was not associated with any patient characteristics such as insurance status or language preference.
The number of clinic visits was the only independent variable in our analysis associated with repeat FOBT adherence. The positive link between clinic visits and CRC screening is well-established [20, 21] and has face validity, since each visit presents an opportunity for staff to assess adherence and, when appropriate, recommend screening. Despite this finding, of the 219 patients who visited EFHC clinics 9–18 months after their initial negative screening, only 31.5 % completed repeat FOBT. The low adherence rate in these patients may be indicative of providers’ inability to consistently deliver preventive care in the face of time pressures and competing clinical demands .
Our findings point to a need for systems-based interventions, implemented outside the realm of typical CHC office visits, to increase repeat FOBT adherence. Several factors suggest the potential for experimentation in this area, including the effectiveness of telephonic  and mail-based  FOBT interventions in non-CHC settings, high acceptability of fecal immunochemical tests [25–27] and new reporting requirements for CRC screening rates at CHCs . If systems-based FOBT programs fail to achieve acceptable levels of repeat annual screening in CHCs, endoscopy may be more effective in reducing CRC mortality in vulnerable populations, assuming it is easier to adhere to a regimen that only requires screening every 5 or 10 years.
This study has several limitations. First, it was conducted at a single urban CHC network whose adult patient population is predominantly uninsured and Spanish-speaking; our findings therefore need to be validated in other vulnerable populations. Second, EFHC staff did not always enter orders in patient charts when distributing FOBT kits. We are therefore unable to determine the extent to which low completion rates were due to providers not offering FOBT versus patients not returning kits distributed by providers. In addition, our exclusion of patients who underwent endoscopy after negative FOBT effectively underestimated overall CRC screening rates, and EFHC’s automated records may not fully capture clinical data for patients who received care outside this CHC network. The small sample size limited statistical power and increased the potential for type II error.
Our use of a 9–18 month screening interval could be considered overly stringent, as selected prior research has found biennial FOBT effective in reducing CRC mortality . We based this follow-up window on USPSTF recommendations, issued in 2008 , which cited approximately equal benefit for annual FOBT and other CRC screening regimens [2, 3]. In order to strike a balance between inclusivity and adherence to the annual screening standard, we allowed patients to complete repeat FOBT up to 6 months after their annual due dates.
Despite these limitations, our study provides important information that should make providers and policymakers question the real-world effectiveness of current FOBT screening practices at CHCs. Observed low repeat FOBT adherence rates—particularly among patients with zero or few visits—highlight the need for new, high-fidelity interventions that do not involve face-to-face interactions between patients and providers. Further research on repeat FOBT among CHC patients, and evaluation of proactive systems-based screening programs, is required to identify strategies with the greatest potential to eliminate disparities in CRC mortality by race/ethnicity [13, 14] and socioeconomic status [15, 16].
Funding support was provided by the Agency for Healthcare Research and Quality (#P01 HS021141). Preliminary findings were presented at the Family Medicine Midwest conference on November 10–11, 2012.