[Skip to Content]
[Skip to Content Landing]
Figure 1.
CONSORT Diagram
CONSORT Diagram

The flow of study patients through randomization, intervention, and outcome analysis.

Figure 2.
Proportion of Overdue Patients Completing Breast, Cervical, and/or Colorectal Cancer Screening During Follow-up: Secondary Outcomes
Proportion of Overdue Patients Completing Breast, Cervical, and/or Colorectal Cancer Screening During Follow-up: Secondary Outcomes

A, Intention-to-treat analyses. P <.01 for all comparisons. B, As-treated analyses. P ≤.002 for all comparisons.

Figure 3.
Rate Differences for All Cancer Screenings
Rate Differences for All Cancer Screenings

Rate differences and 95% CIs for all cancer screenings combined in intervention and comparison groups in patient and practice subgroups.

Table 1.  
Baseline Characteristics Among 1626 Intervention and Control Patients
Baseline Characteristics Among 1626 Intervention and Control Patients
Table 2.  
Mean Cancer Screening Test Completion Rate Over 8 Months of Follow-Up Among 1626 Patients Eligible and Overdue for Cancer Screening at Baseline (Intention-to-Treat and As-Treated Analyses)
Mean Cancer Screening Test Completion Rate Over 8 Months of Follow-Up Among 1626 Patients Eligible and Overdue for Cancer Screening at Baseline (Intention-to-Treat and As-Treated Analyses)
1.
Age-adjusted US death rates and trends for the top 15 cancer sites by race/ethnicity. http://seer.cancer.gov/csr/1975_2012/results_merged/topic_mor_trends.pdf. Accessed October 15, 2015.
2.
Cancer facts & figures for African Americans 2013-2014. Atlanta: American Cancer Society, 2013. http://www.cancer.org/acs/groups/content/@epidemiologysurveilance/documents/document/acspc-036921.pdf. Accessed October 15, 2015.
3.
Braschi  CD, Sly  JR, Singh  S, Villagra  C, Jandorf  L.  Increasing colonoscopy screening for Latino Americans through a patient navigation model: a randomized clinical trial.  J Immigr Minor Health. 2014;16(5):934-940.PubMedGoogle ScholarCrossref
4.
Horne  HN, Phelan-Emrick  DF, Pollack  CE,  et al.  Effect of patient navigation on colorectal cancer screening in a community-based randomized controlled trial of urban African American adults.  Cancer Causes Control. 2015;26(2):239-246.PubMedGoogle ScholarCrossref
5.
Jandorf  L, Braschi  C, Ernstoff  E,  et al.  Culturally targeted patient navigation for increasing african americans’ adherence to screening colonoscopy: a randomized clinical trial.  Cancer Epidemiol Biomarkers Prev. 2013;22(9):1577-1587.PubMedGoogle ScholarCrossref
6.
Lasser  KE, Murillo  J, Lisboa  S,  et al.  Colorectal cancer screening among ethnically diverse, low-income patients: a randomized controlled trial.  Arch Intern Med. 2011;171(10):906-912.PubMedGoogle ScholarCrossref
7.
Percac-Lima  S, Ashburner  JM, Bond  B, Oo  SA, Atlas  SJ.  Decreasing disparities in breast cancer screening in refugee women using culturally tailored patient navigation.  J Gen Intern Med. 2013;28(11):1463-1468.PubMedGoogle ScholarCrossref
8.
Percac-Lima  S, Grant  RW, Green  AR,  et al.  A culturally tailored navigator program for colorectal cancer screening in a community health center: a randomized, controlled trial.  J Gen Intern Med. 2009;24(2):211-217.PubMedGoogle ScholarCrossref
9.
Phillips  CE, Rothstein  JD, Beaver  K, Sherman  BJ, Freund  KM, Battaglia  TA.  Patient navigation to increase mammography screening among inner city women.  J Gen Intern Med. 2011;26(2):123-129.PubMedGoogle ScholarCrossref
10.
Sarfaty  M, Doroshenk  M, Hotz  J,  et al.  Strategies for expanding colorectal cancer screening at community health centers.  CA Cancer J Clin. 2013;63(4):221-231.PubMedGoogle ScholarCrossref
11.
Dohan  D, Schrag  D.  Using navigators to improve care of underserved patients: current practices and approaches.  Cancer. 2005;104(4):848-855.PubMedGoogle ScholarCrossref
12.
Freeman  HP.  The origin, evolution, and principles of patient navigation.  Cancer Epidemiol Biomarkers Prev. 2012;21(10):1614-1617.PubMedGoogle ScholarCrossref
13.
Freeman  HP, Muth  BJ, Kerner  JF.  Expanding access to cancer screening and clinical follow-up among the medically underserved.  Cancer Pract. 1995;3(1):19-30.PubMedGoogle Scholar
14.
Paskett  ED, Harrop  JP, Wells  KJ.  Patient navigation: an update on the state of the science.  CA Cancer J Clin. 2011;61(4):237-249.PubMedGoogle ScholarCrossref
15.
Marshall  JK, Mbah  OM, Ford  JG,  et al.  Effect of patient navigation on breast cancer screening among African American Medicare beneficiaries: a randomized controlled trial.  J Gen Intern Med. 2016;31(1):68-76.PubMedGoogle ScholarCrossref
16.
Percac-Lima  S, López  L, Ashburner  JM, Green  AR, Atlas  SJ.  The longitudinal impact of patient navigation on equity in colorectal cancer screening in a large primary care network.  Cancer. 2014;120(13):2025-2031.PubMedGoogle ScholarCrossref
17.
Rittenhouse  DR, Shortell  SM, Fisher  ES.  Primary care and accountable care--two essential elements of delivery-system reform.  N Engl J Med. 2009;361(24):2301-2303.PubMedGoogle ScholarCrossref
18.
Verma  M, Sarfaty  M, Brooks  D, Wender  RC.  Population-based programs for increasing colorectal cancer screening in the United States.  CA Cancer J Clin. 2015;65(6):497-510.PubMedGoogle ScholarCrossref
19.
Atlas  SJ, Zai  AH, Ashburner  JM,  et al.  Non-visit-based cancer screening using a novel population management system.  J Am Board Fam Med. 2014;27(4):474-485.PubMedGoogle ScholarCrossref
20.
Zai  AH, Kim  S, Kamis  A,  et al.  Applying operations research to optimize a novel population management system for cancer screening.  J Am Med Inform Assoc. 2014;21(e1):e129-e135.PubMedGoogle ScholarCrossref
21.
Berkowitz  SA, Percac-Lima  S, Ashburner  JM,  et al.  Building equity improvement into quality improvement: reducing socioeconomic disparities in colorectal cancer screening as part of population health management.  J Gen Intern Med. 2015;30(7):942-949.PubMedGoogle ScholarCrossref
22.
SRG Technology.  TopCare Patient Population Management Software. Fort Lauderdale, FL: SRG Tech Inc; 2015, http://www.srgtech.com/solutions/topcare/. Accessed November 28, 2015.
23.
U. S. Preventive Services Task Force.  Screening for breast cancer: U.S. Preventive Services Task Force recommendation statement.  Ann Intern Med. 2009. 151(10):716-726, W-236.Google ScholarCrossref
24.
Moyer  V.A., ; U. S. Preventive Services Task Force.  Screening for cervical cancer: U.S. Preventive Services Task Force recommendation statement.  Ann Intern Med. 2012. 156(12):880-891.Google ScholarCrossref
25.
Beaber  EF, Kim  JJ, Schapira  MM,  et al; Population-based Research Optimizing Screening through Personalized Regimens Consortium.  Unifying screening processes within the PROSPR consortium: a conceptual model for breast, cervical, and colorectal cancer screening.  J Natl Cancer Inst. 2015;107(6):djv120.PubMedGoogle ScholarCrossref
26.
Murphy  SN, Chueh  HC.  A security architecture for query tools used to access large biomedical databases.  Proc AMIA Symp. 2002;420(13):552-556.PubMedGoogle Scholar
27.
Atlas  SJ, Grant  RW, Ferris  TG, Chang  Y, Barry  MJ.  Patient-physician connectedness and quality of primary care.  Ann Intern Med. 2009;150(5):325-335.PubMedGoogle ScholarCrossref
28.
Freund  KM, Battaglia  TA, Calhoun  E,  et al; Writing Group of the Patient Navigation Research Program.  Impact of patient navigation on timely cancer care: the Patient Navigation Research Program.  J Natl Cancer Inst. 2014;106(6):dju115.PubMedGoogle ScholarCrossref
29.
Ramachandran  A, Freund  KM, Bak  SM, Heeren  TC, Chen  CA, Battaglia  TA.  Multiple barriers delay care among women with abnormal cancer screening despite patient navigation.  J Womens Health (Larchmt). 2015;24(1):30-36.PubMedGoogle ScholarCrossref
30.
Krok-Schoen  JL, Brewer  BM, Young  GS,  et al.  Participants’ barriers to diagnostic resolution and factors associated with needing patient navigation.  Cancer. 2015;121(16):2757-2764.PubMedGoogle ScholarCrossref
Original Investigation
July 2016

Patient Navigation for Comprehensive Cancer Screening in High-Risk Patients Using a Population-Based Health Information Technology System: A Randomized Clinical Trial

Author Affiliations
  • 1Massachusetts General Hospital, Division of General Medicine, Boston
  • 2Massachusetts General Hospital, Chelsea HealthCare Center, Chelsea
  • 3Harvard Medical School, Boston, Massachusetts
  • 4Massachusetts General Hospital, Laboratory of Computer Sciences, Boston
 

Copyright 2016 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.

JAMA Intern Med. 2016;176(7):930-937. doi:10.1001/jamainternmed.2016.0841
Abstract

Importance  Patient navigation (PN) to improve cancer screening in low-income and racial/ethnic minority populations usually focuses on navigating for single cancers in community health center settings.

Objective  We evaluated PN for breast, cervical, and colorectal cancer screening using a population-based information technology (IT) system within a primary care network.

Design, Setting, and Participants  Randomized clinical trial conducted from April 2014 to December 2014 in 18 practices in an academic primary care network. All patients eligible and overdue for cancer screening were identified and managed using a population-based IT system. Those at high risk for nonadherence with completing screening were identified using an electronic algorithm (language spoken, number of overdue tests, no-show visit history), and randomized to a PN intervention (n = 792) or usual care (n = 820). Navigators used the IT system to track patients, contact them, and provide intense outreach to help them complete cancer screening.

Main Outcomes and Measures  Mean cancer screening test completion rate over 8-month trial for each eligible patient, with all overdue cancer screening tests combined using linear regression models. Secondary outcomes included the proportion of patients completing any and each overdue cancer screening test.

Results  Among 1612 patients (673 men and 975 women; median age, 57 years), baseline patient characteristics were similar among randomized groups. Of 792 intervention patients, patient navigators were unable to reach 151 (19%), deferred 246 (38%) (eg, patient declined, competing comorbidity), and navigated 202 (32%). The mean proportion of patients who were up to date with screening among all overdue screening examinations was higher in the intervention vs the control group for all cancers combined (10.2% vs 6.8%; 95% CI [for the difference], 1.5%-5.2%; P < .001), and for breast (14.7% vs 11.0%; 95% CI, 0.2%-7.3%; P = .04), cervical (11.1% vs 5.7%; 95% CI, 0.8%-5.2%; P = .002), and colon (7.6% vs 4.6%; 95% CI, 0.8%-5.2%; P = .01) cancer compared with control. The proportion of overdue patients who completed any cancer screening during follow-up was higher in the intervention group (25.5% vs 17.0%; 95% CI, 4.7%-12.7%; P < .001). The intervention group had more patients completing screening for breast (23.4% vs 16.6%; 95% CI, 1.8%-12.0%; P = .009), cervical (14.4% vs 8.6%; 95% CI, 1.6%-10.5%; P = .007), and colorectal (13.7% vs 7.0%; 95% CI, 3.2%-10.4%; P < .001) cancer.

Conclusions and Relevance  Patient navigation as part of a population-based IT system significantly increased screening rates for breast, cervical, and colorectal cancer in patients at high risk for nonadherence with testing. Integrating patient navigation into population health management activities for low-income and racial/ethnic minority patients might improve equity of cancer care.

Trial Registration  clinicaltrials.gov Identifier: NCT02553538

Introduction

Despite advances in screening, abnormal test follow-up, and treatment, cancer morbidity and mortality rates for screened cancers are still high, particularly for low-income patients and racial/ethnic minorities.1,2 Patient navigation (PN) has been shown to improve rates of cancer screening in low income and racial/ethnic minority populations.3-10 Patient navigators assist patients in accessing care and guide them through the health care system, helping them overcome identified individual barriers.11-14 Though PN programs usually focus on a single cancer,3-10,15 many patients may be overdue for more than 1 cancer screening test. In addition, most cancer PN programs are located in community health centers or in safety net hospital networks.3-5,8 However, high-risk patients receiving care in settings without on-site PN may not get screened and contribute to persistent cancer care disparities.16

The Affordable Care Act includes provisions that support population health management activities as part of fostering accountable care organizations.17 Population-based programs have been used to increase colorectal cancer (CRC) screening.18 In 2011, a non–visit-based information technology (IT)-enabled population management system was implemented in a large, diverse, academic primary care network to perform preventive cancer screening.19,20 Seeking to improve equity in cancer screening in high-risk patients, an existing PN program in a community health center within the network was expanded to enable navigation of patients at high risk for screening nonadherence in all practices. The PN program was designed to function as a part of the existing population management system.8,21

The objective of our study was to evaluate the impact of population-based, IT-enabled PN for comprehensive breast, cervical, and/or CRC screening in low-income and racial/ethnic minority patients receiving care in a primary care network.

Box Section Ref ID

Key Points

  • Question What effect does patient navigation using a population-based information technology (IT) system within a primary care network have on screening rates for breast, cervical, and colorectal cancer among patients at high risk for non-adherence with screening?

  • Findings The mean proportion of patients who were up to date with screening among all overdue screening examinations was higher in the intervention vs the control group for all cancers combined, suggesting that using patient navigation as part of a population-based IT system significantly increases screening rates for breast, cervical, and colorectal cancer in patients at high risk for nonadherence with testing.

  • Meaning Using patient navigators who are integrated into population health management activities for patients at high risk for screening nonadherence within a practice network may improve equity of cancer care.

Methods
Study Setting

The study was performed from April 10 to December 10, 2014 in the Massachusetts General Primary Care Practice-Based Research Network, consisting of 18 primary care practices, including 4 community health centers. The trial protocol can be found in the Supplement. The network has used TopCare (SRG Technology), a population health IT application, in all primary care practices since 2011.22 TopCare prospectively identifies all patients eligible and overdue for breast, cervical, and/or CRC screening in near real-time. Women 50 to 74 years of age and without prior bilateral mastectomy are eligible for breast cancer screening, and are overdue for screening if they have not had a mammogram or breast magnetic resonance imaging (MRI) conducted in the past 2 years.23 For cervical cancer screening, women without prior total hysterectomy are eligible and overdue if they are 21 to 64 years of age and did not have Papanicolaou testing in the past 3 years, or 30 to 64 years without having undergone Papanicolaou and human papillomavirus (HPV) testing in the past 5 years.24 Patients aged 50 to 75 years without prior total colectomy are eligible for CRC screening and overdue if they did not have a colonoscopy in the past 10 years, or sigmoidoscopy/barium enema/colonography in the past 5 years. Clinicians and staff using the application are presented with real-time, relevant clinical information about their patients and can choose to send reminder letters to patients about screening, use the application to track outreach, or defer screening.19 The Partners network's institutional review board approved all study activities and waived all written informed consent because PN is considered usual clinical care in our institution. Participants were not compensated.

Study Population and Randomization

To identify patients at high risk for not completing screening (nonadherence), we developed an algorithm using pilot data obtained during the first year the IT system was implemented. The algorithm used patient no-show visit history (any no-show visit in the prior year, 1 point), primary language spoken (non-English speaking, 1 point), and number of overdue screening tests (1, 2, or 3 points). Female patients with at least 3 points and male patients with at least 2 points were considered to be at high risk for nonadherence.

We identified 1956 patients at high-risk for nonadherence with cancer screening who were overdue for at least 1 cancer screening test. We excluded 330 patients who received care at the community health center with the existing PN program. Based on our previous PN studies,8 we estimated that over the course of the study period our navigator personnel would be able to reach out to a maximum of 800 patients; therefore, 797 of the remaining 1626 patients were randomized to receive PN, and the rest to usual care. After randomization and prior to the study start date, retrospective review of administrative data excluded 3 patients from the intervention group who left the primary care network, 6 patients who died (3 in each group), and 5 patients less than 50 years old at baseline who were only overdue for breast cancer screening (2 from the intervention group, 3 from the control group). As shown in Figure 1, the final study population consisted of 1612 patients overdue for at least 1 screening at the start of the study period (792 intervention, 820 control). The network's institutional review board approved all study activities and waived all written informed consent because PN is considered usual clinical care in our institution. Participants were not compensated.

Patient Navigation (Intervention Group)

Four part-time, college graduate, patient navigators (2 full-time equivalent), each fluent in 2 to 5 languages, were trained in motivational interviewing, problem solving, goal setting, use of the IT system, electronic medical record documentation, as well as health modules on breast, cervical, and CRC screening. Navigators used in this study had at least 2 years of experience with cancer navigation and continued to work 50% of their time in other established cancer PN programs in our institution. A procedure manual was created to ensure consistency across the navigation team. For each type of screening, the manual outlined phone scripts, talking points, time frame of outreach calls, templates to facilitate documentation in the medical record, and communication with clinicians.

Patients randomized to the intervention group were transferred to a navigator roster within the IT application for the 8-month study period. Navigators used the IT system to track these patients, contact them in their own language, and provide intense outreach to help them complete cancer screening. We used a patient-centered approach. Each navigator was assigned a group of about 200 patients to guide in obtaining screening for all overdue cancer tests. This type of PN is aligned with the conceptual model for breast, cervical, and CRC screening proposed by the Population-based Research Optimizing Screening Through Personalized Regimens (PROSPR) consortium.25 The initial contact, as well as most navigation activities, occurred over the phone. Navigators explored individual barriers to screening, used motivational interviewing to educate and encourage patients, provided reminder calls, arranged transportation, assisted with visit preparation, and accompanied patients to visits if needed. They tracked patients and documented their interventions using case management functionalities from the IT application. In addition, navigators could select appropriate reasons for deferral or exclusion and follow-up on patients with abnormal screening results.

The cost of the initial training, supervision and subsequent study-specific navigator activities was approximately $80 000 for the 8-month study period. The ongoing cost of the established comprehensive cancer screening navigator program is approximately $100 000 per year and reflects primarily personnel cost. The study used the existing network’s population health management IT tool. The costs of maintaining this tool are not included.

Usual Care (Control Group)

Patients randomized to the control group received usual care, which included visit-based reminders within the electronic health record, and non–visit-based outreach by clinicians and staff using the IT application to send patients reminder letters about their overdue cancer screening examinations, call to schedule overdue examinations, or document appropriate reasons for deferral or exclusion.19

Outcome Measures

Patient characteristics and cancer screening data were obtained from an electronic data repository at Partners HealthCare.26 Dates of completion for mammograms, Papanicolaou/HPV tests, and CRC screening tests were obtained from electronic reports or billing data.

The primary outcome was the mean cancer screening test completion rate over the follow-up period for each eligible patient, with all eligible cancers combined in intention-to-treat analyses. The cancer screening test completion rate for each patient was calculated daily, then averaged across the 8-month study period. On any given day, the screening test completion rate was calculated as the number of tests completed divided by the number of eligible tests. We also calculated the completion rate for each individual cancer in the same manner, which translates into the percentage of time screening was up to date during follow-up for each patient. Secondary outcomes included assessing the proportion of patients completing any and each cancer screening during follow-up among those who were eligible and overdue for at least 1 cancer screening at baseline in intention-to-treat analyses. In addition, we conducted as-treated analyses where we removed patients who left our primary care network or who died during follow-up from both intervention and control groups, and also removed patients that the navigators were not able to contact from the intervention group. We compared the primary outcomes in intervention and control patients within relevant subgroups and calculated rate differences and 95% CIs. Patient subgroups were defined by race, primary language, insurance, practice type (community health center vs not), sex, and age.

Statistical Analyses

We used the generalized estimating equations (GEE) approach to take into account the clustering of patient data. The physician was considered the unit of cluster for patients connected with a specific primary care physician (PCP), and the practice was considered the unit of cluster for patients who could be connected with a practice but not a specific PCP.27 For the primary outcome (both intention-to-treat and as-treated analyses), linear regression models with GEE were used to compare the mean completion rate between intervention and control patients for all cancer screening examinations combined and for each individual screening examination. For secondary outcomes (both intention-to-treat and as-treated analyses), logistic regression models with GEE were used to compare the proportion of completed screening tests between intervention and control patients. We conducted a post hoc power analysis. With an intraclass coefficient of 0.044, an average cluster size of 6, and a total of 1612 patients in the intention-to-treat analysis, the design effect was estimated to be 1.2. The study had 80% power to detect a mean difference of 3.1% in mean completion rate for all cancer screening examinations combined. Analyses were conducted using PROC GENMOD, SAS statistical software (version 9.4; SAS Institute).

Results

At baseline, 792 patients were randomized and eligible for the intervention group, and 820 patients were randomized and eligible for the control group (Figure 1). Patient characteristics, including age, sex, primary language, insurance, proportion of patients connected to a specific physician or seen in a community health center, number of clinic visits over the prior 3-years, and risk for nonadherence score were similar between intervention and control groups (Table 1). Among 1612 overdue patients, 46% were overdue for 1 screening, 18% were overdue for 2 screenings, and 36% were overdue for 3 screenings at baseline with a similar distribution among intervention and control groups.

Among 792 intervention patients, 151 (19%) could not be contacted (mean, 3.3 telephone calls per patient). Among the remaining 641 patients, 246 (38%) were deferred for all overdue cancers (eg, patient declined [n = 96], screening completed elsewhere [n = 73], no longer a patient of the practice [n = 40], competing comorbidity [n = 29], anatomically not applicable [n = 5], deceased [n = 3]), and 202 (32%) successfully completed at least 1 overdue cancer screening test (total 252 screening tests; 102 for breast, 57 for cervical, and 93 for colorectal cancer).

Among patients eligible and overdue for cancer screening, the primary outcome, mean cancer-screening completion rates, was higher in the intervention group compared with the control group for all cancers combined and for each individual cancer screening examination (Table 2). In intention-to-treat analyses, intervention patients had a mean cancer screening completion rate 3.4% (95% CI, 1.5%-5.2%; P < .001) higher for all cancer screening examinations combined, and 3.7% (95% CI, 0.2%-7.3%; P = .04) higher for breast, 5.4% (95% CI, 2.1%-9.2%; P = .002) higher for cervical, and 3.0% (95% CI, 0.8%-5.2%; P = .01) higher for colorectal compared with patients in the control group. For colorectal cancer screening, the difference was greater in women (4.6%; 95% CI, 1.7%-7.5%; P = .002) than in men (1.2%; 95% CI, 1.9%-4.2%; P = .46). In as-treated analyses, which removed patients who left the network or who died during follow-up, and who were unable to be reached by the patient navigator during follow-up, the differences between intervention and control groups increased. With all cancers combined, patients in the intervention group had a mean cancer screening rate 5.9% (95% CI, 3.6%-7.9%) higher compared with those in the control group, while it was 7.1% (95% CI, 3.1%-11.2%) higher for breast, 7.9% (95% CI, 3.6%-12.1%) higher for cervical, and 5.0% (95% CI, 2.4%-7.7%) higher for colorectal (P < .001 for all comparisons).

The secondary analyses included the proportion of overdue patients who completed any type of screening during follow-up in intention-to-treat analyses, and was higher in the intervention group compared with the control group (25.5% vs 17.0%; 95% CI [for the difference], 4.7%-12.7%; P < .001) as shown in Figure 2A. In addition, the intervention group had more patients completing screening for breast (23.4% vs 16.6%; 95% CI, 1.8%-12.0%; P = .009), cervical (14.4% vs 8.6%; 95% CI, 1.6%-10.5%; P = .007), and colorectal (13.7% vs 7.0%; 95% CI, 3.2%-10.4%; P < .001) cancer in intention-to-treat analyses (Figure 2A). After excluding patients who left the network or who died during follow-up, or who were unable to be reached by the patient navigator during follow-up, the differences between the intervention and control groups were greater. In these as-treated analyses, 32.9% of intervention patients completed any type of screening during follow-up compared with 18.1% in the control group (difference: 14.8%; 95% CI, 10.3%-19.4%; P < .001). More patients completed breast (29.6% vs 17.6%; 95% CI, 6.2%-17.8%; P < .001), cervical (18.0% vs 9.3%; 95% CI, 3.3%-13.7%; P = .001), and colorectal (18.1% vs 7.6%; 95% CI, 6.3%-14.9%; P < .001) cancer screening in the intervention compared with the control group in as-treated analyses (Figure 2B).

Compared with patients in the control group, patients in the intervention group had a higher rate of screening for all cancers combined in each patient subgroup examined in intention-to-treat and as-treated analyses (Figure 3). The biggest rate differences between patients in the intervention and control groups occurred in patients who were white, spoke English, were seen in community health centers, and were women younger than 50 years.

Discussion

This study evaluated a PN program as a part of visit-independent population health management system for high-risk patients at high-risk for nonadherence with cancer screening in a large primary care practice network. Patients randomized to the PN intervention had significantly higher rates of comprehensive preventive cancer screening compared with patients receiving usual care. The PN intervention improved screening rates among those overdue for breast, cervical, and CRC screening, and the program was beneficial for all high-risk patients regardless of age, sex, insurance, or language spoken.

Similar to previous studies,3-5,7-10 the current study shows that PN significantly improves cancer screening rates in high-risk patients. However, most of these cancer prevention PN studies3-5,7-10 have focused on only 1 single cancer screening. We used a personalized, patient-centered approach to address all overdue tests.25 Each navigator had their own group of patients whom they navigated to obtain screening for all eligible and overdue screening tests.

Patient navigation has been shown to improve cancer care in high-risk populations; however, recent randomized clinical studies highlight that not all high-risk patients need navigation. The greatest benefits were seen in those at high risk of being lost to follow-up.28 Patients with more barriers were less likely to benefit from PN.29 Identifying patients who could most benefit from PN is of great importance when implementing PN programs from a cost-effectiveness perspective. In our primary care network, almost 14 000 patients were overdue for at least 1 cancer screening. We developed an algorithm that enabled us to choose about 11% of patients at high-risk for nonadherence with screening. In our randomized clinical trial, we showed that these patients were unlikely to get screened without PN. Future studies should explore the cost-effectiveness and patient satisfaction with personalized patient-centered cancer screening PN programs.

Although there have been several population-based programs to increase colorectal cancer screening,18 we are not aware of the prior use of a PN program as part of a visit-independent population health management system to identify and navigate patients for comprehensive cancer screening. The IT system enabled us to use an automated algorithm to select patients at high-risk for nonadherence with cancer screening who received care in any network affiliated primary care practice. The navigators used the system to track and navigate patients in real time. Although they were centrally located (in the health center excluded from the study), they were able to navigate patients across all 18 practices. An additional benefit of using the population health management system was that overdue screening tests were scheduled and completed without a clinic visit or clinician involvement (unless a cervical cancer test was done in the primary care physicians’ office). Using a population management, stepped-care approach, the PNs were able to focus on the most high-risk patients in our network; those who had not completed screening, through prior visit-based and population-based efforts.

Limitations of this study include generalizing results to other settings because it was performed in a single academic primary care network with established PN and an advanced IT system. For CRC screening, our network emphasizes colonoscopy, and our results did not include completed home fecal stool tests. The 8-month study period does not exclude the possibility that patients in the control group would have completed tests if followed for a longer time. The automated risk algorithm selected patients at high risk for nonadherence; however, patient navigators were not able to reach 19% of patients randomized to the intervention group. In addition, 38% of these patients had competing comorbidities, obtained screening elsewhere, or declined screenings. Despite the patient navigators being unable to work with a large proportion of intervention patients, our intention-to-treat results showed significantly higher screening rates for each cancer in patients in the PN group. Efforts to improve the risk algorithm by identifying patients most likely to report barriers to cancer care in need of PN,30 or implement a system for providers to directly refer patients to a PN, may further improve the effectiveness of the program.

Conclusions

Patient navigation, using a visit-independent population management system, significantly improved comprehensive cancer screening among patients at high risk of screening nonadherence in a primary care network. Future work should focus on improving the efficiency of patient navigators through better identification of high-risk patients unlikely to undergo screening through routine measures and with barriers that are amenable to PN. Using patient navigators who are integrated into population health management activities for patients at high risk for screening nonadherence within a practice network may improve equity of cancer care.

Back to top
Article Information

Corresponding Author: Sanja Percac-Lima, MD, PhD, Division of General Internal Medicine, Massachusetts General Hospital, 50 Staniford St, Ninth Floor, Boston, MA 02114 (spercaclima@mgh.harvard.edu)

Published Online: June 6, 2016. doi:10.1001/jamainternmed.2016.0841.

Author Contributions: Dr Percac-Lima had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Percac-Lima, Ashburner, Zai, Oo, Guimaraes, Atlas.

Acquisition, analysis, or interpretation of data: Percac-Lima, Ashburner, Zai, Chang, Oo, Atlas.

Drafting of the manuscript: Percac-Lima, Ashburner, Zai, Oo,

Critical revision of the manuscript for important intellectual content: Percac-Lima, Ashburner, Chang, Oo, Guimaraes, Atlas.

Statistical analysis: Ashburner, Chang,

Obtained funding: Percac-Lima, Oo, Atlas.

Administrative, technical, or material support: Ashburner, Zai, Oo, Guimaraes,

Study supervision: Oo, Guimaraes, Atlas.

Other: Zai

Conflict of Interest Disclosures: Massachusetts General Hospital entered into a royalty arrangement on June 27, 2013, to commercialize the population management system with SRG Technology, a for-profit company. Drs Zai and Atlas are beneficiaries of this royalty arrangement but have not received any payments to date. Dr Atlas has received payments as a consultant for the company. Dr Zai has taken a part-time role as chief medical informatics officer for the company. No other disclosures were reported.

Funding/Support: This work was supported by American Cancer Society: Cancer Control Career Development Award for Primary Care Physicians, CCCDAA-14-012-01-CCCDA, Lazarex Cancer Foundation, and the Harvard Medical School Center for Primary Care Innovation Fellowship.

Role of the Funder/Sponsor: The American Cancer Society, Lazarex Cancer Foundation, and the Harvard Medical School Center for Primary Care had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

References
1.
Age-adjusted US death rates and trends for the top 15 cancer sites by race/ethnicity. http://seer.cancer.gov/csr/1975_2012/results_merged/topic_mor_trends.pdf. Accessed October 15, 2015.
2.
Cancer facts & figures for African Americans 2013-2014. Atlanta: American Cancer Society, 2013. http://www.cancer.org/acs/groups/content/@epidemiologysurveilance/documents/document/acspc-036921.pdf. Accessed October 15, 2015.
3.
Braschi  CD, Sly  JR, Singh  S, Villagra  C, Jandorf  L.  Increasing colonoscopy screening for Latino Americans through a patient navigation model: a randomized clinical trial.  J Immigr Minor Health. 2014;16(5):934-940.PubMedGoogle ScholarCrossref
4.
Horne  HN, Phelan-Emrick  DF, Pollack  CE,  et al.  Effect of patient navigation on colorectal cancer screening in a community-based randomized controlled trial of urban African American adults.  Cancer Causes Control. 2015;26(2):239-246.PubMedGoogle ScholarCrossref
5.
Jandorf  L, Braschi  C, Ernstoff  E,  et al.  Culturally targeted patient navigation for increasing african americans’ adherence to screening colonoscopy: a randomized clinical trial.  Cancer Epidemiol Biomarkers Prev. 2013;22(9):1577-1587.PubMedGoogle ScholarCrossref
6.
Lasser  KE, Murillo  J, Lisboa  S,  et al.  Colorectal cancer screening among ethnically diverse, low-income patients: a randomized controlled trial.  Arch Intern Med. 2011;171(10):906-912.PubMedGoogle ScholarCrossref
7.
Percac-Lima  S, Ashburner  JM, Bond  B, Oo  SA, Atlas  SJ.  Decreasing disparities in breast cancer screening in refugee women using culturally tailored patient navigation.  J Gen Intern Med. 2013;28(11):1463-1468.PubMedGoogle ScholarCrossref
8.
Percac-Lima  S, Grant  RW, Green  AR,  et al.  A culturally tailored navigator program for colorectal cancer screening in a community health center: a randomized, controlled trial.  J Gen Intern Med. 2009;24(2):211-217.PubMedGoogle ScholarCrossref
9.
Phillips  CE, Rothstein  JD, Beaver  K, Sherman  BJ, Freund  KM, Battaglia  TA.  Patient navigation to increase mammography screening among inner city women.  J Gen Intern Med. 2011;26(2):123-129.PubMedGoogle ScholarCrossref
10.
Sarfaty  M, Doroshenk  M, Hotz  J,  et al.  Strategies for expanding colorectal cancer screening at community health centers.  CA Cancer J Clin. 2013;63(4):221-231.PubMedGoogle ScholarCrossref
11.
Dohan  D, Schrag  D.  Using navigators to improve care of underserved patients: current practices and approaches.  Cancer. 2005;104(4):848-855.PubMedGoogle ScholarCrossref
12.
Freeman  HP.  The origin, evolution, and principles of patient navigation.  Cancer Epidemiol Biomarkers Prev. 2012;21(10):1614-1617.PubMedGoogle ScholarCrossref
13.
Freeman  HP, Muth  BJ, Kerner  JF.  Expanding access to cancer screening and clinical follow-up among the medically underserved.  Cancer Pract. 1995;3(1):19-30.PubMedGoogle Scholar
14.
Paskett  ED, Harrop  JP, Wells  KJ.  Patient navigation: an update on the state of the science.  CA Cancer J Clin. 2011;61(4):237-249.PubMedGoogle ScholarCrossref
15.
Marshall  JK, Mbah  OM, Ford  JG,  et al.  Effect of patient navigation on breast cancer screening among African American Medicare beneficiaries: a randomized controlled trial.  J Gen Intern Med. 2016;31(1):68-76.PubMedGoogle ScholarCrossref
16.
Percac-Lima  S, López  L, Ashburner  JM, Green  AR, Atlas  SJ.  The longitudinal impact of patient navigation on equity in colorectal cancer screening in a large primary care network.  Cancer. 2014;120(13):2025-2031.PubMedGoogle ScholarCrossref
17.
Rittenhouse  DR, Shortell  SM, Fisher  ES.  Primary care and accountable care--two essential elements of delivery-system reform.  N Engl J Med. 2009;361(24):2301-2303.PubMedGoogle ScholarCrossref
18.
Verma  M, Sarfaty  M, Brooks  D, Wender  RC.  Population-based programs for increasing colorectal cancer screening in the United States.  CA Cancer J Clin. 2015;65(6):497-510.PubMedGoogle ScholarCrossref
19.
Atlas  SJ, Zai  AH, Ashburner  JM,  et al.  Non-visit-based cancer screening using a novel population management system.  J Am Board Fam Med. 2014;27(4):474-485.PubMedGoogle ScholarCrossref
20.
Zai  AH, Kim  S, Kamis  A,  et al.  Applying operations research to optimize a novel population management system for cancer screening.  J Am Med Inform Assoc. 2014;21(e1):e129-e135.PubMedGoogle ScholarCrossref
21.
Berkowitz  SA, Percac-Lima  S, Ashburner  JM,  et al.  Building equity improvement into quality improvement: reducing socioeconomic disparities in colorectal cancer screening as part of population health management.  J Gen Intern Med. 2015;30(7):942-949.PubMedGoogle ScholarCrossref
22.
SRG Technology.  TopCare Patient Population Management Software. Fort Lauderdale, FL: SRG Tech Inc; 2015, http://www.srgtech.com/solutions/topcare/. Accessed November 28, 2015.
23.
U. S. Preventive Services Task Force.  Screening for breast cancer: U.S. Preventive Services Task Force recommendation statement.  Ann Intern Med. 2009. 151(10):716-726, W-236.Google ScholarCrossref
24.
Moyer  V.A., ; U. S. Preventive Services Task Force.  Screening for cervical cancer: U.S. Preventive Services Task Force recommendation statement.  Ann Intern Med. 2012. 156(12):880-891.Google ScholarCrossref
25.
Beaber  EF, Kim  JJ, Schapira  MM,  et al; Population-based Research Optimizing Screening through Personalized Regimens Consortium.  Unifying screening processes within the PROSPR consortium: a conceptual model for breast, cervical, and colorectal cancer screening.  J Natl Cancer Inst. 2015;107(6):djv120.PubMedGoogle ScholarCrossref
26.
Murphy  SN, Chueh  HC.  A security architecture for query tools used to access large biomedical databases.  Proc AMIA Symp. 2002;420(13):552-556.PubMedGoogle Scholar
27.
Atlas  SJ, Grant  RW, Ferris  TG, Chang  Y, Barry  MJ.  Patient-physician connectedness and quality of primary care.  Ann Intern Med. 2009;150(5):325-335.PubMedGoogle ScholarCrossref
28.
Freund  KM, Battaglia  TA, Calhoun  E,  et al; Writing Group of the Patient Navigation Research Program.  Impact of patient navigation on timely cancer care: the Patient Navigation Research Program.  J Natl Cancer Inst. 2014;106(6):dju115.PubMedGoogle ScholarCrossref
29.
Ramachandran  A, Freund  KM, Bak  SM, Heeren  TC, Chen  CA, Battaglia  TA.  Multiple barriers delay care among women with abnormal cancer screening despite patient navigation.  J Womens Health (Larchmt). 2015;24(1):30-36.PubMedGoogle ScholarCrossref
30.
Krok-Schoen  JL, Brewer  BM, Young  GS,  et al.  Participants’ barriers to diagnostic resolution and factors associated with needing patient navigation.  Cancer. 2015;121(16):2757-2764.PubMedGoogle ScholarCrossref
×