Using a comparable continuous control group, our study demonstrates the impact of a patient navigation program in primary care at achieving expected mammography screening rates in a diverse inner-city underserved population. With the exception of Spanish speakers and the Hispanic population, who at baseline had high rates, the navigation intervention increased adherence across all ages, insurance groups, education levels and all other languages and races. Our study design, evaluating a quality improvement navigation program in an entire practice of vulnerable patients, demonstrates the feasibility of adopting this method of care to a clinical setting that mirrors urban safety net settings throughout the country.
Our findings of improved mammogram utilization with a patient navigation intervention are consistent with existing literature15–17
. Dignan and colleagues15
showed improved adherence to mammogram screening over an 18-month period in a randomized controlled trial (RCT) of 157 Native American women utilizing direct patient contact or telephone calls versus a control group. Paskett and colleagues’ RCT16
of 851 subjects also showed improved adherence of mammogram screening in a rural, low income population of white, African American and Native American subjects. Han and colleagues’ study17
utilized patient navigation to improve mammogram adherence of 102 Korean American women after 6 months, but lacked a control group for comparison. Our study findings, designed as a practice improvement within an urban safety-net setting, included a more diverse, yet vulnerable population and thus provide further evidence for the generalizability of navigation as a means to reduce cancer health disparities.
At baseline, more Hispanic and Spanish-speaking women were present in the group allocated to the control group, reflecting that patients are not randomly distributed throughout the practice, but rather are more likely to be seen by Spanish-speaking providers. Prior research has shown both lower27,28
mammography screening rates among Hispanic as compared with white women and may reflect a variety of local factors, including the community’s overall acculturation and education levels, as well as access to insurance and to bilingual health care providers.
Our study was developed to improve HEDIS rates as a quality improvement project within primary care and thus performed as a population-level analysis. The study design responded to a growing emphasis on HEDIS rates as a quality performance measure of individual providers. Interventions like this patient navigation program are becoming increasingly important in order to ensure that practices serving populations with historically lower screening rates achieve benchmarks, both for patient care and for practice reimbursement under pay-for-performance plans. One strength of our study is that it was integrated into the practice with provider “buy-in,” and was designed to evaluate the benefit and effectiveness of integration of this type of program into a busy primary care practice. As such, this model of care fits the Medical Home Model, which is increasingly recognized as a standard way to transform care delivery in primary care settings31–34
We designed our intervention to improve our practice HEDIS measures of biennial screening for all women 51–70 years of age, at a time when the providers were recommending annual screening. Best practice guidelines for patient navigation and timing of the intervention have yet to be defined as reflected by the different intervention protocols implemented in prior studies. Paskett and colleagues16
navigated those ≥12 months overdue for a mammogram and followed them for an additional 12–14 months, while Han and colleagues17
navigated women 2 years overdue and followed up at 6 months. Dignan and colleagues15
navigated for patients 18 months overdue. We chose to initiate navigation after an 18-month screening interval and found this to be an effective strategy in the rational use of the navigator resources, as evidenced by the fact that our greatest effect of the intervention was seen among the groups with the longest time interval since their last mammogram. Even with changes in recent guidelines, our protocol is consistent with recommended mammography frequency in this age group27,35–37
. There is still a subset of women who remain non-adherent to mammography screening using this protocol; the majority of these women were not reachable by phone or mail based on available contact information. This reflects communication challenges in caring for an inner city, at-risk population and suggests a different approach is necessary for this group.
Our study is somewhat limited in its generalizability because it was conducted in only three practices at a single academic safety net institution and required the use of an EMR and information technology support. However, our system reflects standard EMR support and practice systems that have become federal mandates38,39
. Due to limited resources, we were unable to assess costs of the program or patient or provider satisfaction, all of which are crucial to sustainability of such programs.
Our findings support the benefit of patient navigation programs in the primary care setting as one approach to reduce cancer health disparities. While financial support is necessary for primary care providers to develop and maintain such programs, the Medical Home Model31–34
could be one venue to provide the infrastructure and personnel necessary for sustainable navigation implementation. Health care policy-makers should continue to explore advocacy efforts in order to determine how to sustain these programs.