Risk Analysis, Vol. 23, No. 4, 2003

Evaluating Risk Communication about FishConsumption Advisories: Efficacy of a Brochureversus a Classroom Lesson in Spanish and English

Joanna Burger,1∗ Melanie Hughes McDermott,2 Caron Chess,2

Eleanor Bochenek,3 Marla Perez-Lugo,2 and Kerry Kirk Pflugh4

Presentation format can influence the way target audiences understand risk-related informa-tion. Brochures or fish fact sheets are the methods traditionally used by state agencies toinform the public about fish consumption advisories and the risks from consuming fish. Thisstudy examines the efficacy of presenting information about the risks from consuming contam-inated fish and shellfish in two different formats: a brochure and classroom presentation. Thetwo instruments were developed and tested in Spanish and English, reflecting the local ethniccomposition in the Newark Bay Complex. The instruments were tested on women of child-bearing age at the Women, Infants, and Children Center in Elizabeth, New Jersey. Detaileddiagrams were used in both presentations, including contaminated fish species, fish prepara-tion methods, and food chain bioaccumulation and transmission to the fetus. There were fewlanguage-related differences in the efficacy of the classroom lesson, and the main ideas wereunderstood by both groups. Where there were significant differences in understanding aboutthe risks from consuming fish or crabs from the contaminated waters of Newark Bay, in allcases the women exposed to the classroom lesson had a better understanding than those whoread the brochure. Ninety-six percent of the women who heard the lesson understood that itwas unsafe to eat fish from the port, compared to 72% of those reading the brochure. Bothformats succeeded in imparting information to most women about the area under advisories,the fish species under advisories, and transmission of toxins to the fetus. Information on fishpreparation was recalled less clearly, partly because women were asked to relate methodsto reduce the risk from consuming fish from 11 presented, and most recalled only two orthree of the list. The advantages and disadvantages of conducting short classes to women ofchild-bearingage are discussed.

KEY WORDS: Brochure; consumption advisories; fish consumption; language-related differences; lesson;Newark Bay; risk communication

1 Division of Life Sciences, and Environmental and Occupational Health Sciences Institute, Rutgers University, Piscataway, NJ 08854-8082,USA.

2 Center for Environmental Communication, Cook College, Rutgers University, New Brunswick, NJ 08903, USA.3 Haskin Shellfish Research Laboratory, Institute of Coastal and Marine Sciences, Rutgers University, New Brunswick, NJ 08903, USA.4 Chief, Raritan Bureau, Watershed Program, New Jersey Department of Environmental Protection, Trenton, NJ, USA.∗ Address correspondence to Joanna Burger, Division of Life Sciences, 604 Allison Road, Rutgers University, Piscataway, NJ 08854-8082;

burger@biology.rutgers.edu.

791 0272-4332/03/1200-0791$22.00/1 C© 2003 Society for Risk Analysis

792 Burger et al.

1. INTRODUCTION

For many Americans, fishing is an important ac-tivity for food, recreation, and as a way to be out-doors with family and friends. Fish provide manybenefits, both nutritional and social.(1) Fishing is anenjoyable activity that has many social benefits, par-ticularly for Native Americans.(1–4) For some people,fish may be the main affordable source of protein,while for others, it may be the healthiest source. Fishprovide omega-3 fatty acids, which have the potentialto reduce cholesterol levels and reduce the incidenceof heart disease, stroke, and preterm delivery, alongwith other health benefits.(5–10)

In many freshwater and saltwater fish the con-taminant levels are sufficiently high to providepotential adverse health effects, particularly for devel-oping fetuses and young children. A positive relation-ship exists between mercury and PCB levels in fish,fish consumption by pregnant women, and deficitsin neurobehavioral development in children.(11–18)

State agencies have dealt with the potential for ad-verse health effects from eating fish by issuing con-sumption advisories for some waters, as the federalgovernment does not normally do this (except forcommercial fish). Moreover, most state agencies dis-tributed fish consumption guidance mainly with fish-ing licenses (and often no marine fishing licenses arerequired).

Consumption advisories stimulated a flurry ofstudies to determine consumption patterns, the per-ceptions of risk by anglers, and the compliance ofanglers. In general, compliance is not high for manyregions,(19–28) suggesting that risk communication iseither not reaching the target audience, the messageis not believed, or people choose not to follow theadvice.(29–33)

Most state agencies do not have the resources todetermine whether the target audience is receivingthe intended message, whether they understand theconcepts of bioaccumulation and trophic-level trans-fer, and whether some aspects are misunderstood,nor have the economic costs of advisories been de-termined.(34) It is critical to understand how the tar-get audience is understanding the consumption ad-visories or other information provided to it. Whilethere are a great number of studies on the risk fromconsuming fish, on consumption patterns, and on riskperception, there are almost none that evaluate dif-ferent parameters of risk communication or differ-ent risk communication instruments (but see Ref-erences 35 and 36). This is unfortunate, given the

national interest in risk management and commu-nication about fish consumption and consumptionadvisories.(31)

The overall purpose of the project was to designand conduct a community-based education and out-reach program to understand perceptions so that aprogram could be designed to reduce consumptionby Latinos of environmentally contaminated fish andcrabs caught in the Newark Bay Complex, and to pro-vide the communities with information they need tomake informed nutritional choices. In previous workwe showed that knowledge about consumption ad-visories and the risk from consuming contaminatedfish was lowest for Latinos (Hispanics), but when pre-sented with the potential risk they indicated a will-ingness to comply with advisories.(23,37) These studiessuggested a need for an approach directly targetedat the Latino community that emphasizes the rea-sons for limiting consumption. This is particularly im-portant because the Latino population is the fastestgrowing segment of the population in the UnitedStates.(38)

In this article, we examine the efficacy of tworisk communication instruments (a brochure and aclassroom lesson plan) to determine whether subjectswere obtaining the desired messages, and whetherthey would consider changing their fish consumption(or preparation and cooking) behavior. Unlike mostconsumption studies, which survey fishermen (mostof whom are men), our efforts were aimed towardwomen of child-bearing age and pregnant women.The materials were developed in consultation withhealth educators and based on feedback from thewomen themselves. We also compare the responsesof women reading the brochure or attending the classin English and Spanish.

There were several key risk-related concepts thatseemed important to transmit in both instruments:(1) your baby eats what you eat; (2) some fish andcrabs from the Newark Bay Complex are so con-taminated that they should not be eaten by womenof child-bearing age (including pregnant woman);(3) specific fish and crabs should be avoided; (4)there are several methods of preparing and cook-ing fish that reduce the risk from contaminants; and(5) changes in consumption, fish preparation, andcooking methods can reduce risk to unborn chil-dren. Thus the risk communication message wasmore complex than a simple fish consumption ad-visory, and we wished to examine whether the tworisk communication instruments could transmit theseconcepts.

Evaluating Risk Communication about Fish Consumption 793

We also wanted to convey some additional in-formation: (1) some other fish obtained from thesea and sold commercially also contain contaminants(mercury) and should be avoided or intake limited;(2) fish are a healthy and nutritional source of pro-tein; and (3) contaminants increase through the foodchain, and are transmitted from the fish through thewoman to her unborn child. We believed that thelatter point was important because it would providewomen with enough information to understand whylarger fish normally contain higher levels of contami-nants than smaller fish. We aimed to provide sufficientinformation so that women could make their own in-formed decision, and not just avoid fish that couldresult in a decline in overall health.

The Newark Bay Complex includes waters thatflow between New York and New Jersey, and peoplefrom both states fish and crab in the region. It is one ofthe most polluted in the United States,(39–41) and bothNew York and New Jersey issue fish consumptionadvisories.(42,43) Recently it has been reported thatlevels of PCBs in fish from the lower Passaic River,and elsewhere in the New York Bight, exceed the Na-tional Oceanographic and Atmospheric Administra-tion (NOAA) benchmark level.(44,45) Both states issueadvisories for blue crabs (Callinectes sapidus), blue-fish (Pomatomus saltatrix), striped bass (Morone sax-atilis), and American eel (Anguilla rostrata), and NewJersey also has advisories for white perch (Moroneamericana) and white catfish (Ameiurus [Ictalurus]catus). Advisories in the Newark Bay Complex rangefrom do not eat (crab, striped bass), to do not eatfor high risk individuals (e.g., women of child-bearingage and children), and do not eat more than oncea week for the general population (American eel,white catfish, white perch, bluefish over six pounds).New York State recently issued an advisory stat-ing that no women or children should eat stripedbass.(46)

Recently the EPA issued a national advisoryfor freshwater fish because of mercury contami-nation,(47) and the FDA did so for some marinefish.(48) The FDA’s advisory, also based on mer-cury, suggested that pregnant women and women ofchild-bearing age who may become pregnant shouldavoid eating four marine species (or species groups):shark spp., swordfish (Xiphias gladius), king mackerel(Scomberomonus cavalla), and tilefish (Lopholatiluschamaeleonticeps).(48) This is a national problem, as23% of the lakes in the United States, 100% of theGreat Lakes, and 70% of coastal waters are underadvisories.(49)

2. METHODS

2.1. Study Area and Subjects

The Newark Bay Complex of the New York-New Jersey Harbor area was in need of research anddevelopment of outreach targeted to specific sub-populations because of the stringency of fish con-sumption health advisories due to high levels of con-tamination, the increase in the Latino population inthe region, the influx of new immigrants, and previ-ous knowledge of an information gap among ethnicgroups in the region regarding fish contamination.(24)

Instead of concentrating on anglers themselves(a high proportion of whom are men), we derivedour subjects from the Women, Infants, and ChildrenCenter (WIC) in Elizabeth, New Jersey (Fig. 1).

WIC provides federal grants from the U.S. De-partment of Agriculture to states for supplementalfoods, health care referrals, and nutritional educa-tion for low-income pregnant, breastfeeding, and non-breastfeeding postpartum women, and to infants andchildren found to be nutritionally at risk.(50) Thesecenters are aimed at improving women’s health andnutrition, particularly during pregnancy and the child-bearing years. Thus, it is a federal program that couldbe applicable to all states. The program serves about7 million people each month.(50) There are federal in-come eligibility guidelines: a woman in a family offour must have an annual family income of less than$33,485.

Interviewing was completed from January 14through March 1, 2002, following an initial round ofpretesting in the fall of 2001.

2.2. Development of Brochure and ClassroomLesson Plan

The overall protocol was to develop risk com-munication materials aimed at informing pregnantwomen and women of child-bearing age about therisks from consuming contaminated fish from theNewark Bay Complex. We made the choice of formatfor risk communication materials based on reviews ofmany fish fact sheets and fish consumption advisories,and consultation with community leaders and organi-zations, public health providers, and nutrition educa-tors. While there are state fishing advisory brochures,they do not contain information that is specifically tar-geted to pregnant women, or on ways to reduce riskthat are aimed at a Hispanic audience.

Focus groups and interviews held with healthand nutrition managers and staff pointed to face-to-

794 Burger et al.

Fig. 1. Map of the Newark Bay Complex showing the advisory area covered by the brochure and classroom lesson, and the location of theWomen, Infants, and Children Center in Elizabeth, New Jersey.

face communication, such as an interactive lesson,as the most effective form of outreach. In additionto a lesson plan, we chose to develop a brochuredue to the low cost of production and ease of widedistribution. Both instruments were entitled “Yourbaby eats the fish you eat!”(51) We aimed for a fifth-grade reading level, and used extensive graphics andother features of presentation found to be optimalfor audiences with low levels of English language lit-eracy.(35,52) The brochure was a single page folded toprovide six different panels, and the classroom lessonconsisted of a teacher’s guide and 44 slides in Power-Point format (for a class of 15–25 minutes duration).The slides, which can be displayed in PowerPoint asoverheads or on a flipchart, were designed to be self-explanatory.

After each instrument was developed, it waspretested in interviews for the brochure (N = 43),and in two focus groups with educators for the lessonplan (N = 8, 13). Each was then modified based onresponses and suggestions received before the finalproducts were tested again to determine the efficacyof each in conveying the risk message (brochure: N =46, survey: N = 51). The tables present survey results

for the second round of testing on the revised versionof the brochure and lesson plan.

After the initial brochure and lesson plan weredeveloped in English, they were translated intoSpanish by M. Perez-Lugo, as were all subsequentversions (Fig. 2). Subjects could choose the languagethey wished to read; the English version was evalu-ated in English, and the Spanish version was eval-uated in Spanish. This process resulted in a finalbrochure and classroom lesson plan that could be usedstatewide.

2.3. Testing of Instruments

Each day we determined on a random basis whichinstrument we were going to test (brochure or class-room), and then all women present were approachedand asked to participate. Only two women refused,although a few left before the interviews were com-pleted (due to being called to the doctor office).Whenever possible, women were given a choice oflanguage. In some cases, out of preference or if thetiming was more convenient, immigrants fromSpanish-speaking countries elected to review English

Evaluating Risk Communication about Fish Consumption 795

Fig. 2. Diagrams of food chain concepts used in the brochure andin the lesson plan.

materials. Because many of the women had smallchildren with them, we provided a selection of smallpresents, including coloring books, activity books,stickers, and crayons, as a thank you (which also aidedin keeping the children occupied during interviews).

As would be expected, different approaches wereused to introduce the women to the brochure and les-son plan. However, regardless of whether they readthe brochure or listened to the lesson plan, the ques-tions we asked of them were the same. Questions wereadministered in one-to-one interviews. Women wereapproached as they sat in the waiting rooms, and askedif they would volunteer to look at the brochure (orattend the classroom lesson) and answer a few ques-tions. The classroom lesson was taught during regu-lar, weekly sessions required by the WIC program.Women who volunteered to stay on to be interviewed

were thanked with a gift certificate to a local depart-ment store.

Many of the questions could be answered by yes,no, or I do not know, or other categorical answers.Other questions were open-ended, and we scored re-sponses as relevant (according to preset categories),or as correct, liberally interpreted. For example, onthe question about the area covered, we counted cor-rect responses ranging from “the Newark Bay” to “theport” (of Elizabeth). In the tables we give a short formof the question actually asked of the women.

2.4. Statistical Evaluation

In the text and tables we give the percent ofpeople giving a “yes” response, responses equiva-lent to our preset categories, or the correct response.Statistics, however, were computed on the raw data.We used SPSS (Statistical Package for the SocialSciences) Pearson chi-square tests, which give achi-square value and level of significance. If expectedvalue for a cell was less than 5, the analysis was notconsidered significant.

One open-ended question (What are some of theways you can make sure the fish you eat are safe?) didnot have mutually exclusive answers, making a chi-square test inappropriate. Thus for this question weused a binomial test to test the hypothesis that peoplewho got the lesson took home more of the messagesthan people who only read the brochure.

3. RESULTS

3.1. Demographics

The mean age of women interviewed was 28 yearsfor the brochure and 29 years for the classroom les-son, and most people had one to three children. Wedid not ask educational and income information be-cause this would have reduced participation in thesurvey; additionally, WIC has an income ceiling, en-suring that the program is reaching only low-incomewomen and children (by definition of the U.S. Depart-ment of Agriculture).(50)

Only 7% (brochure) to 19% (English classroom)of the women were pregnant at the time (Table I).Only 36% of women who read the brochure and28% of those who attended the class were born inthe United States; most of the rest were born inPuerto Rico or in other Latin American countries.Over 90% of both samples consumed fish, although amuch smaller group admitted to eating fish or crabsfrom the Newark Bay Complex.

796 Burger et al.

Table I. Demographics of Women Sampled in Interviews on theEfficacy of Using the Brochure or a Classroom Lesson on theRisks from Fish Consumption. English and Spanish Refer to

Their Choice of Language for the Brochure or Class

ClassroomBrochure

(English and Spanish) English Spanish

Sample size 45 26 25Mean age (±SD) 28 ± 7.0 26 ± 11.2 32 ± 8.5

in yearsMean number of 2.0 ± 1.3 1.3 ± 0.7 2.4 ± 1.3

children (±SD)Percent pregnant 7 19 12Percent that consume 91 100 88

fishPlace of birth

U.S. 16 14 1Mexico 2 0 4West Indies 9 5 6Central America 7 1 6South America 11 3 8Europe 0 1 0Africa 0 2 0

3.2. The Classroom Lesson Plan

The classroom lesson proved to be an effec-tive method of imparting a number of complicatedconcepts. In general, there were few differences asa function of whether they listened to the lesson inEnglish or Spanish (Table II); this was also the casewith the brochure (unpublished data). A number ofwomen noted that they were pleased that the classwas taught in Spanish, and that this was a first forthem. One hundred percent of respondents correctlyrecalled that the species under advisory were unsafe,and correctly identified the area of concern as “theport” or other areas within the Newark Bay Com-plex. All but one woman in each class stated that shebelieved the warning about the fish under advisory,that eating fish caught in the Port of Elizabeth wasnot a good idea (or words to that effect), and that apregnant woman could harm her baby by doing so(Table II).

One of the objectives of the risk communicationinstruments was to impart information on ways to re-duce the risk from contaminants in fish. We expecteda lower correct response rate on the question about“ways to make sure that the fish you eat are safe” be-cause there were 11 correct methods discussed in theclass, and any given person might be expected to re-call only some of them. Nonetheless, some methodswere more readily recalled than others, including re-moving fat, cooking on a rack, and eating fish that are

purchased, rather than caught by family or friends(Table II). There were some small differences be-tween the responses to the English versus the Spanishversion, including removing fat and buying fish ratherthan eating fish they catch (Table II). Some responsessuggested that information about contaminated fishfrom our classroom lesson had not been transmitted,and people retained concepts based on previous un-derstanding about fish, including “look at the eyes andgills,” “soak in lemon juice,” and “smell it.”

One of the more complex concepts discussed wasthe movement of contaminants from small fish to largefish, from fish in the water to fish on the plate, andfrom the fish to the mother to the fetus (Fig. 3). Mostwomen had a generally correct understanding of themovement of chemicals up the food chain, althoughit was rarely expressed in this manner. For example,the answers we accepted in response to the question“Can you describe what you think this picture [Fig. 3]is trying to show?” included “big fish eat small fish,”“the chain of food,” “the food chain,” and “the bigfish has more contamination than the small fish, andthat ends up on your plate.” Others understood theconcept completely, and said “The bigger fish eats thesmaller fish. The bigger fish is more contaminated, andyou eat the bigger fish” or “The bigger fish eats thesmaller fish and that is the one that is contaminatedon the plate.”

Two other questions were particularly of interest:“Imagine you were talking to a pregnant friend aboutthe information in this pamphlet, what would you tellher?” and “What is the main advice this pamphletgives here?” (Table III). Although most women wouldgive their friends correct advice, some did not relaythe correct message (see the last 3 in Table III).

Another significant finding is that a full 70% ofrespondents stated that they intended to change howthey chose or prepare fish they ate as a result of whatthey learned from the lesson. If we exclude those whosaid they do not eat fish, that figure rises to 74%.Moreover, 90% of the women who say that a familymember fishes (10% of the total) indicated an intentto make these changes in behavior. This latter cate-gory represents the most vulnerable group and thelesson’s key target audience.

3.3. Comparison of Brochure and ClassroomLesson Plan

An important objective of our research was todetermine whether a classroom lesson with a teacherwas more effective than having women read only abrochure (Table IV). Where there were significant

Evaluating Risk Communication about Fish Consumption 797

Table II. Differences in the Efficacy of a Classroom Lesson in Spanish and English Concerning Risk and Fish Consumption

Question English Spanish X2(p)

Sample size 26 25Does anyone in your family catch fish or crabs? (yes) 26 12 1.8 (NS)Was the class clear and easy to understand? 100 100What do you think about eating fish from the port? (given is the percent who gave the following

response)It is not a good idea. 96 96 2.0 (NS)

Do you think that a pregnant woman can harm her baby by eating fish from the port?(yes response) 96 100 0.9 (NS)

What did the instructor say about the six types of fish?Unsafe 100 100 0.5 (NS)Contaminated 42 64Do not eat 77 84

What are some of the ways you can make sure the fish youa eat are safe? (percent who mentionedthis method)Eat safe fish 0 4Avoid fish from Newark Bay, species under advisory 12 16Remove fat 62 32Cook on a rack 54 72Throw away juices 8 24Do not fry or coat the fish 31 32Buy fish (rather than eat fish you catch) 35 8Don’t buy or eat swordfish, etc. 8 8Eat other fish only up to 12 oz. a week 0 4Ask where fish came from 23 16Eat smaller fish 23 8Vary types of fish 4 4 (NS)ac

Do you believe fish from Newark Bay came from an unsafe location? (yes) 96 96 1.1 (NS)What is the area the class refers to? (percent who gave the following response)

Newark Bay 100 100 0.0 (NS)What is the picture with the red dots in fish showing? (percent giving correct response)

contaminant movement, food chain, fish are contaminated92 80

What is the picture with the red dots in the pregnant woman showing? (percent giving correctresponse) contaminants moving into woman and baby

96 96 0.0 (NS)

Do you plan to change the fish you eat or how you cook them? (yes) 64 76 0.9 (NS)

aYou or your family.bThese are the six fish with health advisories.cBinomial test.Notes: Given are percentages. For most questions the correct answer is yes and that is the percent given, unless otherwise noted. NS = notsignificant.

differences in understanding about the risks fromconsuming fish or crabs, most woman exposed to theclassroom lesson had a better understanding thanthose who read the brochure. That is, for 18 of 20 ques-tions, those hearing the lesson did better than thosereading only the brochure (X2 = 165.5, p < 0.001).Ninety-six percent of the woman who heard the lessonunderstood that it was unsafe to eat fish from the port,compared to 72% of those reading the brochure. Ev-ery woman who listened to the lesson understood thatthe six species of fish and crabs were unsafe, whereas91% of those who read the brochure did. Nonetheless,both formats succeeded in imparting information tomost women about the area under advisories, the fish

species under advisories, and bioaccumulation up thefood chain.

4. DISCUSSION

4.1. Brochure and Lesson Development

Two important aspects of this study that we feelcontributed to its success were: (1) close consultationwith community-based organizations and with healthcare and nutrition education officials and staffers, and(2) an interactive process of outreach material de-velopment involving an interactive process of con-sultation, testing, and revision (Fig. 2). The former

798 Burger et al.

Fig. 3. The process of developing the brochure and lesson planinvolved an interactive process of several stakeholders.

is essential because these constituents provided valu-able advice about the development of the two instru-ments. Moreover, it encouraged buy-in from healthcare management and staff, making them more likelyto distribute brochures and teach the lessons (whichwere designed to be given by health educators withno prior knowledge of the subject). Indeed, the NewJersey State WIC program has approved the les-son plan and brochure for use at all WIC sites inthe Newark Bay area. As an extension of this re-search project, we are providing training and mate-rials for the use and distribution of the lesson planat these sites. While numerous brochures preparedby outside groups are distributed at WIC sites, it isperhaps unlikely that the state agency would haveadopted the lesson plan as part of its own curricu-lum if the management and staff (at both the lo-cal Elizabeth and state levels) had not participatedin the review and development of the materials. In-cluding stakeholders in decision making has becomea popular concept, both for management and forresearch.(53–55)

Interactive testing of outreach materials (riskcommunication instruments) with the intended au-

Table III. Representative Quotes from Woman whenAsked:“What Advice Would You Give a Pregnant Friend?”

Don’t eat fish from the port.Fish is a good source of protein but she should be careful what fish

she eats. She should ask where fish came from and she shouldremove the head, guts, skin and fat.

Don’t eat bluefish, shark, blue crabs. I’d say don’t eat any fishfrom Bay.

She has to know where fish is coming from. But she should eat alot of fish because it is good for the baby.

I’d tell her that she should eat fish because it is good for her andthe baby but she should buy it in the supermarket and askwhere it came from.

Don’t eat fish unless bought in supermarket.Before buying fish ask where it comes from because the fish from

port is dangerous. Too much bacteria.To be careful with the fish she buys and eats.Be careful what fish you eat. Ask where it comes from.I’d give her the pamphlet you gave us.Don’t eat any fish caught from all areas of New Jersey.a

To eat fish.b

I wouldn’t eat the fish if you are pregnant because it is dirty.a

aThis was obviously not our message.bLacks information on the advisory.

dience, as well as continuous interaction with peo-ple and agencies ultimately responsible for their dis-tribution is also essential because it allows refine-ment of the instruments for specific local conditions.This approach is about more than “reaching thetarget audience”; it is about continuously involvingthem and those who communicate health informa-tion directly to them in the development of materi-als that will provide them with the information theyneed to make informed decisions about their familieshealth.

While the actual length of the classroom lessonwas longer than the brochure, and the time devotedto each differed, the same material was presented inboth. The difference was that in the classroom les-son, each idea was presented in more depth and high-lighted.

4.2. Universality of Results

Interest in examining the brochure and in listen-ing to the classroom lesson was relatively high, per-haps because it provided information directly of inter-est to women concerning the health of their children.Over 90% of the women approached agreed to par-ticipate, a clear indication of interest, or at least open-ness. Although there are few other studies with whichto compare these findings, Burger and Waishwell(36)

Evaluating Risk Communication about Fish Consumption 799

Table IV. Efficacy of Using a Brochure versus a Class Lesson Concerning the Risk from Consumption of Fish

Brochure Lesson X2(p)

Sample size 45 51Does anyone in your family catch fish or crabs? (yes) 15 20 0.3 (NS)What do you think about eating fish from the port? (given is the percent who gave the following response)

It is not a good idea. 72 96 11.6 (0.003)Do you think a pregnant woman can harm her baby by eating fish from the port?

(yes response) 89 98 3.9 (NS)What did instructor say about the six species of fish? (percent who said the following)

Unsafe 91 100 4.6 (0.03)What are some of the ways you can make sure the fish you eat are safe? (percent who said each of the

following)Eat safe fish you catch 4 2Avoid fish from Newark Bay, species under advisory 9 14Remove fat 46 47Cook on rack 39 62Throw away juices 13 15Don’t fry or coat the fish 13 31Buy fish (rather than eat fish you catch) 20 21Don’t buy or eat swordfish, etc. 0 8Eat other fish only up to 12 oz. a week 2 16Ask where fish came from 17 22Eat smaller fish 2 15Vary types of fish 0 4 (0.0001)a

Do you believe fish from Newark Bay came from an unsafe location? (yes) 87 98 4.7 (NS)What is the area the class referrs to? (percent giving the following response) 86 100 0.8 (NS)

Newark BayWhat is the picture with the red dots in fish showing? (percent giving the correct response) (contaminant

movement, food chain, fish are contaminated)72 86 2.9 (NS)

What is the picture with red dots in pregnant women showing? (percent giving the correct response)(contaminant moving into fish, woman, and baby)

100 96 0.8 (NS)

Do you plan to change the fish you eat and how you cook them? (yes) 62 70 0.6 (NS)

aBinomial test.Note: Percentages are given for the answer in parenthesis. NS = not significant.

found a similarly high interest rate in people (bothmen and women) fishing along the Savannah Riverin South Carolina. One aspect that seemed im-portant was the direct approach by enthusiasticresearchers.

While we tested the efficacy of two differentmethods of risk communication about fish consump-tion and found a high level of knowledge was obtainedfrom both instruments, the results may not be as pos-itive when brochures are not presented in person andwhen the classroom lesson is given by a less enthusi-astic and informed teacher. In both cases, the instru-ments were administered in the subject’s preferredlanguage of choice by enthusiastic people. Further,the brochure and class were presented in a women’sclinic where interest in infant and child health washigh. The relatively high success of communicatingthe message under these circumstances suggests thatwomen’s health clinics and similar locations should bespecifically targeted.

While the data presented in this article are spe-cific for the New York-New Jersey harbor area, theymay provide useful insights on risk communication forLatinas, for women who consume fish, and for agen-cies developing informational brochures and otheroutreach materials for pregnant women, women withyoung children, and women in their child-bearingyears. Information on the study will be availablethrough traditional scientific journals, but also via theWeb.

4.3. Efficacy of a Brochure versusa Classroom Lesson

Where there were significant differences in un-derstanding about the risks from consuming fish orcrabs from the contaminated waters of Newark Bay,the woman exposed to the classroom lesson had a bet-ter understanding than those who read the brochure.

800 Burger et al.

This result may be a function of several factors, includ-ing additional time devoted to the problem (it tookless time to read the brochure than to listen to the15–25 minute presentation), the use of multiplemodalities all reinforcing the same message (oralcommentary, written text, and visual images), theengagement of interest by a “live” presenter, andfinally, the interactive format in which participantswere asked questions, volunteered information, andhad the opportunity to question the instructor. Fur-thermore, the advantages of the lesson over thebrochure may be higher under real-life circumstances,under which a woman who is handed a brochure maynot take the time to read it at all.

The disadvantage of classroom presentation in-clude finding a suitable venue for the classes, and re-cruiting audiences and informed and enthusiastic in-structors. Further, it is far more expensive to conductclassroom lessons than to print brochures. Brochurescan be disseminated to far more places, and can bepicked up by any interested parties.

Optimally, both formats should be available to thepublic. The lesson plan provides an in-depth discus-sion of the problem, where questions can be asked.The brochure provides a written explanation thatcan be taken home, referred to, and given to friendsand family. Moreover, brochures could be distributedwith fishing licenses, as well as at maternity clinics,health clinics, and child care facilities. Connelly andKnuth(35) similarly found that multiple formats maybe required to meet the needs of anglers and fishconsumers.

4.4. Risk Management

In many European communities, consumers ac-cept co-responsibility for solving environmental prob-lems by changing their consumption patterns.(56) Inthe past, environmental problems were handled pri-marily by regulating the main sources of pollution, butrecently consumers are being asked to take responsi-bility to avoid unsafe products.(57) This charge oftenleads to confusion, partly because of lack of knowl-edge, and to noncompliance. In dealing with contam-ination in fish from lakes, rivers, and estuarine wa-ters, state agencies, as well as the U.S. EnvironmentalProtection Agency and the Federal Drug Administra-tion, have relied on the issuance of fish consumptionadvisories. Government agencies and others may pre-fer to follow a strategy of relying on advisories sincethat provides a means of reducing adverse health con-sequences while avoiding potentially large clean-up

costs.(34) However, when governments take this ap-proach, it may be ineffective in protecting the healthof the public.

There are several risk management problems in-herent in using fish consumption advisories in thismanner:(28) (1) knowledge about advisories is oftenlimited, (2) correct knowledge about the complexi-ties of which fish to eat, how much to eat, and thepopulation at risk is limited, (3) even with such knowl-edge, compliance is often low, and (4) the highest-riskgroup among the target audience (pregnant womenand those of child-bearing age) often do not receiveword of them (in most states they are distributed withfishing licenses). The problem is further complicatedfor estuarine waters because most states do not havea marine fishing license that can serve as a conduit fordissemination of brochures.

Low compliance may be the result of ignoringthe underlying reasons for fishing. One assumptionof advisories is that people go fishing to catch fish toeat, and if they only understood that some fish arecontaminated and can cause adverse health effects(to unborn and young children), they would ceasethis behavior. This is a naive assumption. Toth andBrown(1) clearly showed that there are a myriad of so-cial and economic reasons that people fish, and thatwithin many ethnic communities fishing plays a keysocial role. In a recent study of people fishing in theNewark Bay Complex, Burger(58) asked why they an-gled, using both open-ended questions and a ratingscale. Relaxation, being outdoors, and outdoor sportsranked the highest, followed by getting away fromdemands and communing with nature. Addressingthese needs seems critical for changing consumptionpatterns.

Rather than aim risk communication exclusivelyat the angling public, another way of approachingthe problem is to target the population at risk di-rectly (namely, women of child-bearing age, thosepregnant and breast feeding in particular, and chil-dren). Presenting them with relevant information onboth the benefits and the adverse effects of consumingfish allows them to make an informed decision aboutnot only eating fish, but what fish to eat, obtainedfrom what source. Fish consumption is by no meansa simple problem, since a consumer must evaluatethe health and economic benefits of eating fish (andself-caught fish), the disadvantages of eating contam-inated fish, and the relative merits of specific species,obtained from specific locations, prepared and cookedin different manners. This is a complicated decisionprocess.

Evaluating Risk Communication about Fish Consumption 801

The results of this study suggest lessons thatcan be examined further for this and otherareas: (1) collaboration with health care and nutritioneducation agencies and staff can increase understand-ing and buy-in by those who disseminate information;(2) iteration in the development and testing of ma-terials is critical to producing materials relevant for,and comprehensible by, the target audience; (3) pro-vision of information in the appropriate language isessential; (4) presentation of materials directly to thetarget audience and to those responsible for and in-terested in fetal and infant health gets information tothose who need it most and whose behavioral changeswill make the greatest impact; (5) presentation of out-reach materials by enthusiastic people is likely to in-crease effectiveness; and (6) balancing the approachis essential to achieving outcomes that maximize goodhealth (fish are health-enhancing food for these rea-sons, but can be unsafe for particular people for otherreasons).

Finally, the brochure and lesson plans developedas part of this study, and freely available on the In-ternet (http://aesop.rutgers.edu/cec/fish/html), can bemodified and tested in other communities with otherethnic groups. While the species of fish to avoid mayvary, other aspects, such as preparation and cookingmethods, will not. The process used to initially de-velop the materials can be usefully applied to modifythese materials for local conditions. That is, consult-ing with health officials and educators, and testing thematerials with the target audience, will ensure modi-fications that apply locally.

ACKNOWLEDGMENTS

We wish to thank the administrators and staff ofthe Women, Infants, and Children Center in Elizabethfor their extensive collaboration and willingness to al-low the use of their facility for testing, K. Ensle andthe staff of Rutgers Cooperative Extension of UnionCounty for participating in focus group testing andproviding other helpful advice, B. Ruppell for infor-mation on chemical contamination and fish consump-tion advisories in Newark Bay, D. Riley on healthimplications, E. Bordel for illustrations and graphicdesign, numerous agencies and organization in UnionCounty for consultation and advice, and membersof the multi-agency Toxics in Biota Committee fortechnical advice and review. Helpful comments onthe manuscript were provided by M. Gochfeld. Pri-mary funding was provided by the Division of Science,Research and Technology of the NJ Department of

Environmental Protection. Additional funding wasprovided by the New Jersey Agricultural ExperimentStation, Center for Environmental Communication,Environmental and Occupational Health Sciences In-stitute, NIEHS (ESO 5022, JB), Union County Boardof Chosen Freeholders, and Consortium for RiskEvaluation with Stakeholder Participation (DOE,DE-FG 26-00NT 40938). The results, conclusions, andinterpretations reported herein are the sole responsi-bility of the authors, and should not in any way beinterpreted as representing the views of the fundingagencies.

REFERENCES

1. Toth, J. F., Jr., & Brown, R. B. (1997). Racial and gender mean-ings of why people participate in recreational fishing. LeisureSciences, 19, 129–146.

2. Harris, S. G., & Harper, B. L. (1997). A Native Americanexposure scenario. Risk Analysis, 17, 789–795.

3. Egeland, G. M., & Middaugh, J. P. (1997). Balancing fish con-sumption benefits with mercury exposure. Science, 278, 1904–1905.

4. Burger, J. (1999). American Indians, hunting and fishing rates,risk and the Idaho National Engineering and EnvironmentalLaboratory. Environmental Research, 80, 317–329.

5. Hunter, D. J., Kazda, I., Chockalingam, A., & Fodor, J. G.(1988). Fish consumption and cardiovascular mortality inCanada: An inter-regional comparison. American Journal Pre-ventive Medicine, 4, 5–11.

6. Horn, E. (1992). Toxics in seafood. Tidal Exchange, 3, 6–7.7. Anderson, P. D., & Wiener, J. B. (1995). Eating fish. In J. D.

Graham & J. B. Wiener (Eds.), Risk versus Risk: Tradeoffs inProtecting Health and the Environment (pp. 104–123). Cam-bridge, MA: Harvard University Press.

8. Iso, H., & Rexrode, K. M. (2001). Intake of fish and omega-3fatty acids and risk of stroke in women. Journal of AmericanMedical Association, 285, 304–312.

9. Nestel, P. (2001). Effect of fish oil and fish on cardiovasculardiseases. Current Artherosceral Research, 3, 68–73.

10. Olsen, S. F., & Secher, N. J. (2002). Low consumption ofseafood in early pregnancy as a risk factor for preterm de-livery: Prospective cohort study. British Medical Journal, 324,447–450.

11. Jacobson, J. L., Humphrey, H. E. B., Jacobson, S. W., Schwartz,S. L., Mullin, M. D., & Welch, R. (1989). Determinants ofpolychlorinated biphenyls (PCBs), polybrominated biphenyls(PBBs), and dichlorodiphenyl trichloroethane (DDT) levelsin the sera of young children. American Journal of PublicHealth, 79, 1410–1404.

12. Jacobson, J. L., Jacobson, S. W., & Humphrey, J. B. (1990).Effects of in utero exposure to polychlorinated biphenyls andrelated contaminants on cognitive functioning in young chil-dren. Journal of Pediatrics, 116, 38–45.

13. Institute of Medicine. (1996). Seafood Safety. Washington,DC: National Academy Press.

14. Sparks, P., & Shepherd, R. (1994). Public perceptions of thepotential hazards associated with food production: An empir-ical study. Risk Analysis, 14, 799–808.

15. Agency for Toxic Substances and Disease Registry (ATSDR).(1996). States issue a record number of health advisories. Haz-ardous Substances Public Health, 6, 1–2.

802 Burger et al.

16. Jacobson, J. L., & Jacobson, S. W. (1996). Intellectual im-pairment in children exposed to polychlorinated biphenyls inutero. New England Journal of Medicine, 335, 783–789.

17. Schantz, S. L. (1996). Developmental neurotoxicity of PCBsin humans: What do we know and where do we go from here?Neurotoxicology and Teratology, 18, 217–227.

18. Weihe P., Grandjean, P., Debes, F., & White, R. (1996). Healthimplications for Faroe Islanders of heavy metals and PCBsfrom pilot whales. Science for the Total Environment, 186, 141–148.

19. Reinert, R. E., Knuth, B. A., Kamrin, M. A., & Stober, Q. J.(1991). Risk assessment, risk management, and fish consump-tion advisories in the United States. Fisheries, 16, 5–12.

20. Burger, J., & Gochfeld, M. (1991). Fishing a Superfund site:Dissonance and risk perception of environmental hazards byfishermen in Puerto Rico. Risk Analysis, 11, 269–277.

21. Burger, J., Cooper, K., & Gochfeld, M. (1992). Exposure as-sessment for heavy metal ingestion from a sport fish in PuertoRico: Estimating risk for local fishermen. Journal of Toxicol-ogy and Environmental Health, 36, 355–365.

22. Burger, J., Staine, K., & Gochfeld, M. (1993). Fishing in con-taminated waters: Knowledge and risk perception of hazardsby fishermen in New York City. Journal of Toxicology andEnvironmental Health, 9, 95–105.

23. Burger, J., Pflugh, K. K., Lurig, L., vonHagen, L. A., & von-Hagen, S. (1999). Fishing in urban New Jersey: Ethnicity af-fects information sources, perception, and compliance. RiskAnalysis, 19, 217–229.

24. Burger, J., Stephens, W., Boring, C. S., Kuklinski, M., Gibbons,J. W., & Gochfeld, M. (1999). Ethnicity and risk: Fishing andconsumption in people fishing along the Savannah River. RiskAnalysis, 19, 427–438.

25. Velicer, C. M., & Knuth, B. A. (1994). Communicating con-taminant risks from sport-caught fish: The importance of tar-get audience assessment. Risk Analysis, 14, 833–841.

26. May, H., & Burger, J. (1996). Fishing in a polluted estuary:Fishing behavior, fish consumption, and potential risk. RiskAnalysis, 16, 459–471.

27. Knuth, B. A. (1995). Fish consumption health advisories: Whoheeds the advice. Great Lakes Research Reviews, 1, 36–40.

28. Burger, J. (2000). Consumption advisories and compliance:The fishing public and the deamplification of risk. Journal ofEnvironmental Planning and Management, 43, 471–488.

29. Belton, T., Roundy, R., & Weinstein, N. (1986). Urban fisher-men: Managing the risks of toxic exposure. Environment, 28,19–37.

30. Fiore, B. J., Anderson, H. A., Hanrahan, L. P., Olson, L. J., &Sonzogni, W. G. (1989). Sport fish consumption and body bur-den levels of chlorinated hydrocarbons: A study of Wisconsinanglers. Archives of Environmental Health, 44, 82–88.

31. Environmental Protection Agency (EPA). (1998). AssessingHuman Health Risks from Chemically Contaminated Fish andShellfish: A Guidance Manual. EPA-503/8-89-002, AppendixF. U.S. Environmental Protection Agency, Cincinnati, OH.

32. Ebert, E. S. (1996). Fish consumption and human health: De-veloping partnerships between risk assessors and resourcemanagers. American Fisheries Society Symposium, 16, 261–170.

33. Reinert, R. E., Knuth, B. A., Kamrin, M. A., & Stober, Q. J.(1996). Risk assessment, risk management, and fish consump-tion advisories in the United States. Fisheries, 16, 5–12.

34. Jakus, P. M., Dadakas, D., & Fly, J. M. (1998). Fish con-sumption advisories: Incorporating angler-specific knowl-edge, habits, and catch rates in a site choice model. AmericanJournal of Agriculture Economics, 80, 1019–1024.

35. Connelly, N. A., & Knuth, B. A. (1998). Evaluating risk com-munication: Examining target audience perceptions about

four presentation formats for fish consumption health advi-sory information. Risk Analysis, 18, 649–659.

36. Burger, J., & Waishwell, L. (2001). Are we reaching the targetaudience? Evaluation of a fish fact sheet. Science of the TotalEnvironment, 227, 77–86.

37. Pflugh, K. K., Lurig, L., vonHagen, L. A., vonHagen, S., &Burger, J. (1999). Urban anglers’ perception of risk from con-taminated fish. Science of the Total Environment, 228, 203–219.

38. Huerta, E. E., & Macario, E. (1999). Communicating healthrisk to ethnic groups: Reaching Hispanics as a case study. Jour-nal of the National Cancer Institute, Monograph, 35, 23–26.

39. Ayres, R. U., & Rod, S. R. (1986). Patterns of pollution in theHudson-Raritan Basin. Environment, 28, 14–43.

40. O’Connor, T. P., & Ehler, C. N. (1991). Results from theNOAA National Status and Trends Program on distributionand effects of chemical contamination in the coastal and es-tuarine United States. Environmental Monitoring and Assess-ment, 17, 33–49.

41. Squibb, K. S. (1992). Overview of toxics in the harbor estuary.Tidal Exchange, 3, 1–2.

42. New York State Department of Health (NYSDOH). (1994).1994–1995 Health Advisories: Chemicals in Sportfish or Game.New York State Department of Health, Bureau of Toxic Sub-stance Assessment, NY.

43. New Jersey Department of Environmental Protection(NJDEP). (1994). A Guide to Health Advisories for EatingFish and Crabs in New Jersey. New Jersey Department of En-vironmental Protection, Trenton, NJ.

44. Finley, B. L., Trowbridge, K. R., Burton, S., Proctor, D. M.,Panko, J. M., & Paustenbach, D. J. (1997). Preliminary as-sessment of PCB risks to human and ecological health in theLower Passaic River. Journal of Toxicology and Environmen-tal Health, 52, 95–118.

45. Deshpande, A. D., Draxler, A. F. J., Zdanowicz, V. S., Schrock,M. E., & Paulson, A. J. (2002). Contaminant levels in the mus-cle of four species of fish important to the recreational fisheryof the New York Bight Apex. Marine Pollution Bulletin, 44,164–177.

46. New York State Department of Health (NYDOH). (2002).2000–2001 Health Advisories: Chemicals in Sportfish andGame. New York State Department of Health, Bureau ofToxic Substance Assessment, NY.

47. Environmental Protection Agency (EPA). (2001). Consump-tion Advisories. Environmental Protection Agency, Office ofWater. Available at www.epa.gove/ost/fishadvice/advice.html.

48. Food and Drug Administration (FDA). (2001). FDA Con-sumer Advisory. Available at http//www.fda.gov/bbs/topics/ANSWERS/2000/advisory.html.

49. Environmental Protection Agency (EPA). (2001). Update:National Listing of Fish and Wildlife Advisories. Environmen-tal Protection Agency, Office of Water. EPA-823-F-01-010,Washington, DC.

50. Woman, Infants, and Children (WIC). (2002). Availablehttp://www.fns.usda.gov/wic/about WIC.htm.

51. McDermott, M. H., Chess, C., Kirk Pflugh, K., Perez-Lugo, M.,Bochenek, E., & Burger, J. (2003). Communicating a complexmessage to the population most at risk: An outreach strategyfor fish consumption advisories. Journal of Applied Environ-mental Education and Communication, 2, 39–48.

52. National Institutes of Health. (2001). Clear and Simple: De-veloping Effective Print Materials for Low-Literate Readers.Washington, DC: National Cancer Institute, National Insti-tutes of Health.

53. Commission on Risk Assessment and Risk Management.(1997). Risk Assessment and Management in RegulatoryDecision-Making. Washington, DC: U.S. Government.

Evaluating Risk Communication about Fish Consumption 803

54. National Research Council (NRC). (1994). BuildingConsensus through Risk Assessment and Managementof the Department of Energy’s Environmental Reme-diation Program. Washington, DC: National AcademyPress.

55. Siegrist, M., & Cvetkovich, G. (2000). Perception of hazards:The role of social trust and knowledge. Risk Analysis, 20, 713–719.

56. Halkier, B. (1999). Consequences of the politicization of con-

sumption: The example of environmentally friendly consump-tion practices. Journal of Environmental Policy & Planning, 1,25–41.

57. Eden, S. E. (1993). Individual environmental responsibilityand its role in public environmentalism. Environment andPlanning, 25, 1743–1758.

58. Burger, J. (in press). Fish consumption advisories: Knowledge,compliance and why people fish in an urban estuary. Journalof Risk Research.