Addressing many, speaking to one

Editor’s note: Steve Appel is president of Appel Research, LLC, Niskayuna, N.Y. Barbara Bird, RNC of Bird Consulting Group, a Clifton Park, N.J., health care consultancy, The project described in this article was done in conjunction with Lisa Grace of Conceptual Images for the Bureau of Ambulatory Care Services, AIDS Institute, New York State Department of Health (Roberta Glaros, director), and largely funded by the Centers for Disease Control and Prevention. Appel and the other moderators involved, including Francesca Moscatelli and Nancy Banks, are Qualitative Research Consultants Association members.

Statistics show that infant mortality rates in parts of New York State rank near those of many third-world countries. Since there is a clear link between proper prenatal medical care and positive pregnancy outcomes, the obvious solution is to convince more women to get prenatal care.

That’s exactly what the AIDS Institute, a division of the New York State Department of Health, had in mind when it hired a team of consultants * a health care expert, a marketing researcher and a media producer - to develop a multimedia campaign to address this mission.

The task was to get the hardest-to-reach pregnant women to seek early prenatal care. To succeed, we would have to convince these women - often drug addicts, convicted felons, illegal aliens, refugees - to tell us what it would take to win them over.

The method would be marketing research, using focus groups and Perception Analyzers. A key use of those little black boxes with antennae was trade-off analysis (the ability to measure messages and images in comparison to each other) - in addition to the more common second-by-second graphing of radio and TV spots. The major benefit of the Perception Analyzer system is that you get not only the yes or no opinions but also a measure of their intensity.

Unique aspects

There were two unique aspects to this project for the Department:

  • The possible messages and images were developed and produced and then, through this technology, were tested on separate targeted audiences before the roll-out.
  • A less direct approach of outreach via the media was chosen, in contrast to the common practice in social marketing, which holds that effective convincing is done face-to-face.

The Department of Health has made great strides using the paradigm of individual outreach through community-based organizations (CBOs). The results have been positive, with several neighborhoods showing marked improvements. In areas such as central Harlem, the infant mortality rate is now just about equal to the average for the entire United States.

Yet there remains a hardcore population of women who neither seek nor receive prenatal care and who are therefore more likely to have high-risk pregnancies. They have proven to be hard to persuade on an individual basis - assuming you can locate them. It was time for a new approach - using a multimedia campaign to change behaviors.

These hardest-to-reach are relatively few, but costly in every sense of the word. They are costly in terms of human suffering and they are costly to taxpayers, who pay, through Medicaid, hundreds of thousands of dollars in medical care in an attempt to get low-birthweight babies through their early weeks and months.

Target areas

The target areas were drawn from ZIP codes in New York City that contain high rates of late or no prenatal care among women giving birth, and of women giving birth who were HIV-positive. For our focus groups, we recruited African-American women from east Harlem, Spanish-speaking women from the west Bronx, African-American teens from the Bedford-Stuyvesant section of Brooklyn, and Haitian-Creole-speaking women from Brooklyn’s Crown Heights and Flatbush sections.

All had to be pregnant currently or within the past five years, with at least one of the following characteristics: homeless, a survivor of domestic violence, a substance user within past five years, had been incarcerated within past five years, or had exchanged sex for drugs or money in past five years. We didn’t ask if the subjects were HIV-positive or an undocumented immigrant, although it was often known by the recruiters, who were from CBOs that provided services to the populations we were studying. Keeping with the theme of contacting only the hardest-to-reach women, the organizations were instructed to recruit women they had been unable to serve.

Just as the recruiting was anything but traditional, so too were the rooms we used to hold the focus groups. We felt it was important to conduct our research in a location in which the women were familiar and comfortable, but there are no professional facilities in any of these neighborhoods. A representative of a medical school clearly didn’t get it when we asked for help in finding a facility that had an observation space with a one-way mirror. She suggested that we use a police precinct interview room. (Not a Law & Order viewer, we surmise.)

Recognizing the challenges faced in gathering data from respondents who were not expected to be forthcoming, along with special cultural concerns, we used qualified moderators who were sensitive to the characteristics of the focus group respondents. We also employed simultaneous translators, so the team could keep track of the proceedings in real time.

Not all the advice we received was accurate. With regard to cultural issues, we were told Haitians especially would not participate in front of a video camera. We found little difficulty - the Haitians were just as forthcoming as the other groups.

In fact, we were surprised by the openness exhibited by all the respondents. Without exception, they were extremely enthusiastic about the process, largely because they felt it was rare for someone to legitimately solicit their opinions. The result, time and time again, was a motivated focus group. Moreover, they were so oblivious to the videotaping that several pulled their shirts up to rub their pregnant bellies, even with a man in the room operating the Perception Analyzer system.

First round: messages and images

Two rounds of focus groups were conducted. The first round was relatively conventional. The women were encouraged to discuss their previous experiences with prenatal care and benchmark their attitudes about the process, the barriers, and their decision-making in obtaining prenatal care, if they had sought any.

Among their reasons for not getting prenatal care, they said that a medical exam would reveal drug use, thereby proving the mother unfit. The result would be that her other children would be taken away. We were also told of “alternative” methods of health care. “When I’m pregnant, I fast every Tuesday until I deliver,” said a Haitian woman. Another said, “I go to a herbalist, a doctor of leaves.”

To test the emotional reactions to colors and designs that might be used in the awareness campaign, we handed out fabric swatches. There was a clear favorite among most of the respondents, a fabric consisting of a dark blue sky, with clouds, stars, the moon, and the sun with happy faces. “It makes you, like, happy, or think of babies and stuff. Because I would like to put this...in the kids’ room.” It was also described as cute, soft, and a reminder of heaven, “where the angels are.” The teens tended to choose the more child-like, as opposed to baby-like, fabrics.

The respondents in the first round were asked to draw two pictures, one of women who had had prenatal care, and a second of women who hadn’t. From the drawings, the message was clear: there are distinct physical and psychological differences between the two types of women.

We found that the overall theme presented was the “good girl/bad girl” dichotomy. The woman who received prenatal care was described as happy, knowing that she was taking the correct action for the benefit of herself and her baby. The woman who did not seek medical care was seen as sad, lacking in self-esteem, the object of pity, and one who cares far more about herself than about her baby. The Spanish-speaking respondents were especially judgmental, showering praise on the “good girl,” while calling the “bad girl” selfish and irresponsible.

From a consensus among respondents, five possible messages were derived from the first round:

1) You owe yourself good care, and we want to help you get it.

2) Protect yourself and your baby. Get prenatal care.

3) Together, we’re there for you and your baby.

4) A big step toward a healthy start...early prenatal care.

5) As soon as I think I’m pregnant, I’ll do the right thing and get prenatal care.

These messages were all positive statements. The negative approaches — the idea that either 1) failure to get pregnancy care will result in an unhealthy baby, or 2) that the mother risked having her baby taken from her by the authorities - were viewed as ineffective in promoting early prenatal care.

Another message that was rejected was the notion of “no hassles.” “Nothing in New York has no hassles,” stated one respondent. Furthermore, pregnancy itself is a hassle. As evidence of discordance in such a message, one teen said, “If you wasn’t going through no hassle, your baby’s not going to be healthy.”

In addition to the overall theme, we determined that certain elements, or facts, had to be included in the message. For example, it was important to state that health care for pregnancy is free or low-cost, that the care is confidential, and that there is a toll-free telephone number to get information.

We also considered the question of whose images should be included in the messages. Should it be the women, the baby, the doctor, or some combination of them? The conclusion was that, since the messages were all aimed at the pregnant woman, she should be featured.

The groups also discussed the use of celebrities as spokespersons and rejected that notion, arguing that only a real person who had had similar experiences would be believed. “They [the actors] are going to be fake, taking us for suckers.”

Second round: testing the print and electronic media before the roll-out

The second round of discussions featured the opinions and reactions of a different group of similarly qualified women. From this data and from consultation with clients, a series of messages, images (drawings and photographs), and audio/video spots were advanced for testing.

Print
The print themes and images were presented in various combinations, using the Perception Analyzer trade-off feature as well as discussion. In addition, the respondents listened to and viewed audio and video spots, using the Perception Analyzer moment-to-moment function to determine which they thought would have the greatest impact on women like themselves.

For the print phase, the evaluation process was started by measuring each of the five messages individually.

The respondents were instructed to set their Perception Analyzer dials to 50, a neutral position halfway between 0 and 100. We then presented Message A and asked them to move their dials to a higher number if they found the message persuasive (100 would represent perfection), or to a lower number if the message was not persuasive (0 would represent a total lack of persuasion). We continued the process with the other four messages. We found that we could not combine the scores from all the groups because, culturally, each group was unique in their starting points. For example, by the numbers, the Spanish-speaking respondents were the most optimistic about the efficacy of these messages, averaging 87 for the five, which does not leave much room for variation, from top to bottom. In contrast, the teens were not so generally convinced. Their scores for the five messages averaged only 64.

This preceding exercise could easily have been done on paper, but not so the next step: ranking the five messages via the trade-off feature. The respondents were again asked to set their dials at 50 and compare one message to another. The moderator began by holding up a sign with Message A (“You owe yourself good care and we want to help you get it.”) in her left hand and Message B (“Protect yourself and your baby. Get prenatal care.”) in her right hand. The respondents were instructed to turn the dial to the left, toward zero, if they favored Message A; Message B would be represented by turning the dial to the right, toward 100. If the respondent favored Message A to the total exclusion of the other message, then they would turn their dials all the way to zero. If they had a slight preference, then they might move their dials to 45, or perhaps 55 if they favored Message B. With five messages to compare to each other, there were 10 permutations.

We then showed five images - photographs on backgrounds - whose concepts and colors had been derived from the first round. The colors incorporated into the images were the two most favored: blues and purples. The photographic inclusion of clouds into the images of three women with and without umbrellas came from the response to the overwhelming favorite fabric pattern tested in the first phase, one of a blue background with clouds, suns, stars and moons.

Using the same system, we ranked the five images:

1) three women holding a translucent sphere

2) one woman holding a translucent sphere

3) three hands, two cupping around the third

4) three women holding two umbrellas

5) three women without umbrellas.

In the observation room during one of these sessions, Bureau of HIV Ambulatory Care Services Director Roberta Glaros commented, “I’m typically skeptical of technology used in this field because social science research tends to focus on questions that have quantifiable answers. The Perception Analyzer allows measurement of a person’s qualitative judgment. It really worked.”

The methodology enabled us to measure not only the favored choice among the five in each category, but also the intensity of the favorable feelings toward the winning message or image, measured against each of the other choices. We do want to note, however, that we used the information gathered during the discussion following each of the numerical rankings to reinforce the confidence we had in the final selections.

Figure 1

The winning messages and images varied among the different ages and cultures. The African-American adults selected “Protect yourself and your baby. Get prenatal care,” overlaid on the image of the three women with umbrellas (seen as a “sisterhood of safety”).

The African-American teens favored the same message but with a different image - three hands, two cupping around the third.

The Spanish-speaking women chose “You owe yourself good care and we want to help you get it,” superimposed on the photograph of the three hands, two cupping around the third. After language consultation, this was modified to read, in Spanish, “You deserve good prenatal care and we want to help you get it.”

The Haitians selected the message, “Protect yourself and your baby. Get prenatal care,” in Haitian-Creole, along with the picture of the three women with umbrellas. Initially, the image of the three hands led, but in the trade-off, the three women with umbrellas prevailed. The umbrella was viewed as a protector of fragile pregnant women.

Radio and TV
For the radio and TV commercials, the respondents held the Perception Analyzer dials as they viewed or listened to the 30-second spots, each with themes, messages and images developed from the research results, reflected in the language of the respondent. The words and pictures were measured on a second-to-second basis, through the visual use of a line graph overlaid on the video image, or for the audio, overlaid on a timeline. We measured each spot twice, early in the focus group, and then 90 minutes later toward the end, in order to simulate the real-life experience of repetition and measure how effective each would be after hearing it or seeing it several times.

Here is a script used for a 30-second spot that tested well, using the actual words from one of the respondents:

“During your pregnancy, you got to really look at, let me go check myself out. If my head ain’t together, my body ain’t gonna come along. First of all, I let them know they really don’t have anything to fear. Fear is one of the things that keep people away from going to the clinic. Scared, not scared. High, not high. Because if you can go out there and get other things that you need that’s not really good for you, then you could go out there and get something that you need that’s going to help you.”

The campaign

The multimedia campaign began in the targeted ZIP codes in December 2001. A telephone hotline was established, with operators trained to counsel and refer. It was started with print materials - billboards, posters, bus shelter signs, bus ads, pamphlets and stickers. Radio and some television spots began airing after a few months. In November 2002, 30-second videos were placed in movie theatres, interspersed among the previews for coming attractions.

Meanwhile, the New York State’s Bureau of Child Health, another Department of Health division charged with moving more women to prenatal care, piggybacked onto our campaign, taking the print ads statewide, using a different toll-free phone number.

Did it work?

Two measures are being used to evaluate the effectiveness of this multimedia campaign. First, the number of hotline calls is being tallied. Second, records are being kept at locations where women are directed for prenatal care to determine how the women were motivated to come in.

Currently, the hotline is receiving several hundred calls per month. Generally, the number of calls has increased during periods when radio spots are broadcast. Twice as many callers said they heard about the campaign on radio than from any other source, followed by posters and printed materials. We recognize, of course, that there is a synergy between the various points of contact, and that the radio ads are pushing the women across the threshold that had been built by the print ads.

The least likely source of information was community outreach, a key finding. Apparently, we are succeeding in our objective - to get to women who had not been reachable through conventional methods in social marketing.

As a side benefit, a few dozen calls per month were received that were for other services not specifically related to prenatal care. These were referred to appropriate agencies.

While it is too early to tell if infant mortality rates have dropped in the targeted areas, this remains the ultimate goal, along with fewer problem pregnancies.

And as researchers, we were gratified to note the concluding sentiment from several of the respondents: “Thank you for listening to me.”