Cancer screening rates among Latinas are suboptimal. Compared with whites, Latinas have lower screening rates for breast (mammogram within past year, 46.5% vs 51.5%), cervical (74.7% vs79.1%), and colorectal cancer (CRC) (51.2% vs 60.2%) (1). Improving screening rates is important for reducing racial/ethnic disparities in rates of cancer. However, screening can be burdensome, particularly for women who are age-eligible for multiple screenings. Adding to this burden is that different screenings (ie, breast, cervical, CRC) have different recommended screening intervals and that recommendations vary among organizations and have changed over time (2–5). These variations may cause confusion about when screening should commence or how often they should take place (6–8), and, coupled with barriers to screening, may contribute to disparities in screening rates.

Health care organizations often remind patients of overdue screenings via letters or telephone. Such reminders can increase screening rates (9,10), but a reminder system may be difficult to sustain in community health centers (CHCs) that provide care to low-income populations and have limited resources. Interactive voice response (IVR) messages — automated telephone calls that allow recipients to interact with a prerecorded message through speaking or pushing telephone keys — are a promising low-cost strategy for patient reminders. The technology allows participants to hear an initial message and then additional menu options as desired. The decision to access menu options is driven by the recipient, and staff involvement is minimal once the IVR programming is developed. IVR messages can be tailored by language, name, and other variables. Messages are easily updated, can be delivered at designated times, and are inexpensive to deliver. Importantly, IVR messages are viewed favorably in a medical context (11). Although IVR messages have had mixed success in promoting cancer screenings (12–15), success with this technology could increase if more information about how to design IVR messages to meet recipients’ needs was available (16). Thus, we conducted a series of focus groups with Latinas in need of cancer screening(s) to 1) explore perceptions of and barriers to cancer screening and 2) understand audience communication norms related to the design of IVR messages.

We conducted 7 focus groups in Spanish with Latina patients in need of 1 or more cancer screenings. Focus groups were held in August and September 2008. Five groups were conducted with women needing 1 screening (n = 28) and 2 groups with women needing multiple screenings (n = 12). This design allowed us to examine differences in emergent themes between participants in need of 1 or multiple screenings. All participants were patients of a CHC in a city in northeast Massachusetts that is predominately Latino (>75%) (17) and has high rates of poverty and limited resources (18, 19). Approximately 80% of Latinos living in the city receive care at the CHC; most of the CHC patients are Hispanic (90%) and are of Caribbean origin (ie, Dominican Republic, Puerto Rico) (18).

All study participants (n = 40) self-identified as Latina, mean age was 51.9 (range, 22–86), and 12 (30.0%) needed multiple screens (Table 1). Nineteen (47.5%) participants had less than a high school education. Thirty-six participants (90.0%) reported having a physician or nurse practitioner at the CHC whom they viewed as their regular health care provider. Although all participants reported the ability to converse in Spanish, 7 (17.5%) participants preferred to speak in English.

We successfully recruited Latinas in need of cancer screening into our qualitative study and found that most participants in our focus groups were receptive to the idea of receiving multiple IVR messages to prompt scheduling of needed screenings. Our findings indicate that messages tailored to a Latina audience would have more value than generic messages and that messages identifying the screenings needed and addressing certain concerns may be important to encourage Latinas to seek screening.

Other research shows that Latino patients will access health education information via telephone (26). Thus, IVR messages are likely to be an acceptable strategy to promote the scheduling of needed screenings and initiate the screening process. Although the barriers identified in our study are generally in line with barriers reported elsewhere (27,28), it was important to obtain this information from our target population to inform the design and tailoring of the IVR messages.

Our findings can guide the development of IVR messages to promote cancer screening among Latinas. First, IVR messages should be delivered by a native speaker from the region represented by the target population. Using a locally recognized voice (eg, person who records CHC’s telephone recordings) and a customary greeting would likely increase the level of call acceptability and meaningfulness. Second, the telephone number that appears in caller ID should identify the CHC; this may reduce the number of hang-ups. Third, messages should specify the date of the patient’s last screening and clearly state that her physician, EMR, or other CHC records indicate screening is overdue. Many study participants incorrectly believed they were current with needed screenings. Because screening recommendations change and such changes create confusion about when to get tested (6,7), CHC’s practice guidelines should also be stated. Fourth, messages should 1) emphasize the preventive role of screening (to alleviate concerns that the call is relaying information about a health problem), 2) recognize and address the fear and embarrassment often associated with screenings, and 3) include the idea of family as a motivator for screening. For some participants in our study, a lack of insurance or inadequate insurance prevented screening; screening rates among women who have insurance are higher than among women who do not (29,30). Thus, IVR messages should include an option that allows recipients to connect with CHC staff to discuss insurance-related questions.

Tailoring IVR messages and identifying all screenings needed in 1 message may be better strategies than creating generic messages and identifying each screening needed in separate messages, especially because the additional costs of following these strategies would be nominal. Fear of screening procedures was a barrier identified by all focus groups but was especially relevant for women needing multiple screenings. Messages could be crafted to let recipients know that although multiple screenings are needed, they can be completed at different times.

Focus group participants, especially those needing a single screening, expressed fear of screening outcomes, and this fear prevented or delayed screening. This finding reinforces the importance of IVR messages that focus on the long-term, beneficial outcomes (ie, early detection may lead to better prognosis) of completing screenings. Our study also highlights the importance of addressing provider-level factors in IVR messages (eg, setting up a referral system for needed screenings) and encouraging patients to take action to get screened. Some participants were waiting for their provider to recommend screening; IVR messages should encourage patients to discuss screening with their providers. IVR messages alone are likely to have a limited effect on cancer screening rates among CHC patients unless they are implemented with other systems-level efforts. Educational campaigns such as bilingual flyers posted in the examination room are easy to implement and may be an effective way to reach patients. In addition to simple efforts, more comprehensive (and likely more effective) strategies should be considered: for example, clinic-supported systems-level interventions such as tailored EMR templates and provider education.

This study has several limitations. Because our study was conducted to inform the development of an intervention to be implemented at 1 CHC, we recruited participants only from this site. Although the CHC serves the entire city, it serves a primarily Latino population of Caribbean origin, thus limiting the generalizability of the study results. Although all participants were able to converse in Spanish, 7 participants preferred to speak English; those who preferred English may be more acculturated than those who preferred Spanish. We did not assess level of acculturation or explore how it may affect personal views about cancer screenings. Women who participated in our study may have been different from women in need of cancer screening(s) who elected not to participate. Finally, we could not access EMRs to determine whether participants who reported that they had not received a provider referral (needed to begin the colonoscopy process) were accurate in their recall.

Given that prompting methods such as letters and telephone calls made by staff are often difficult to sustain, IVR messages hold promise as a feasible, cost-effective strategy to promote cancer screening among underserved Latina patients. As CHCs continue to develop information technology systems, opportunities exist to increase the systematization and efficiency of e-health (eg, IVR messages, text message) strategies to remind patients of overdue cancer screenings that are both sustainable and scalable. These efforts can reduce organizational burden cost-effectively, especially if the EMR can automatically initiate IVR calls according to individual need. Reducing the amount of staff time needed to confirm screening needs and initiate call reminders would free up time to focus outreach efforts on patients who do not respond to automated messages. Future research is needed to test the effectiveness of IVR messages in promoting the scheduling and completing of cancer screenings.