To examine the extent of parental knowledge about lead poisoning and its prevention and to determine characteristics associated with accurate lead knowledge.
Twenty-three pediatric practices and 1 family practicein Chicago, Ill, and its suburbs.
A 24-question test regarding lead poisoning and its prevention (Chicago Lead Knowledge Test) was developed based on lead specialists' review and parental test-retest reliability. One point was assigned for each correct response. It was self-administered by a sample of 2225 parents of 0- to 6-year-old children visiting study practices. A 1-way analysis of variance (ANOVA) was used to determine the association of demographic descriptors with test scores.
Respondents had a mean age of 33 years. Ninety percent were mothers, 49% were college graduates, and 80% were home owners. Fifteen percent lived in homes built before 1950, of which 36% were remodeled or renovated during the last 6 months. Respondents' youngestn children were 80% white, 10% Hispanic, 5% African American, and 5% other. Ten percent received Medicaid and 86% had other medical insurance. Thirty-four percent recalled receipt of lead information from a health care provider, and 2.4% had had a child with a blood lead level of 0.48 µmol/L (10 µg/dL) or higher. The mean Chicago Lead Knowledge Test score was 12.2 (SD, 3.7). Questions related to lead exposure were more often answered correctly than those related to prevention and diet. In the ANOVA model, those who recalled receipt of lead information from a health care provider, college graduates, respondents aged 30 years or older, Hispanic respondents, and those living in homes built before 1950 had higher scores (all ANOVA P≤.001).
Parents do not have much knowledge of ways to prevent childhood lead poisoning. Information from a health care provider can aid parental knowledge. The Chicago Lead Knowledge Test is a new self-administered tool to help evaluate lead education programs.
LEAD POISONING continues to be a common problem in some urban areas. Chicago, Ill, is one such area, where 10.9% of children tested in 1996 (12,881/118,156) were newly identified with a blood lead level of 0.72 µmol/L (15 µg/dL) or higher.1 The resources in Chicago to abate lead hazards are limited. Therefore, the major focus of lead poisoning prevention programs is often parental education. Such programs merit evaluation concerning their effectiveness.
An important evaluation technique for an educational program is the measurement of parental knowledge. This requires the creation of a reliable measurement tool that assesses the array of topics one would convey in an educational program. This study was designed to (1) develop a test of lead knowledge that could be used in a variety of settings to evaluate lead education programs, (2) use it to describe parental knowledge about lead, and (3) examine parental sociodemographic characteristics and effect of health care provider education on accurate parental lead knowledge. To accomplish these aims, the Chicago Lead Knowledge Test was developed and included in a survey of parents in a variety of physicians' offices throughout Chicago and its suburbs.
An initial set of test questions was developed during the evaluation of the Chicago Neighborhood-Based Childhood Lead Primary Prevention Project, a jointly funded lead education and prevention project of the Centers for Disease Control and Prevention and US Department of Housing and Urban Development. The questions were modified after review by lead experts, including physicians, nurses, and social workers at Children's Memorial Hospital, Chicago; health care providers attending a regional meeting of the Chicago Area Health Care Providers' Lead Consortium; and Illinois and Chicago public health lead program leaders.
The test-retest reliability of the instrument was assessed on a convenience sample of consenting parents who were enrolled at the emergency department, inpatient floors, and clinics of Children's Memorial Hospital. The parents completed the survey either by themselves or with a research assistant reading the questions if the parent so requested. Within 2 weeks, an identical survey was administered to all parents surveyed the first time. Parents of hospitalized children completed the second survey in the same manner as the one originally administered. The other parents were surveyed by telephone. The final test instrument (Chicago Lead Knowledge Test) included 24 questions, each with "true," "false," or "don't know" response choices.
Data were analyzed using SPSS for Windows, release 6.1.2 (SPSS Inc, Chicago, Ill). Phase 1 analyses included calculation of test-retest agreement and computation of the Pearson correlation between test scores (the number of correct responses) on the 2 occasions.
The anonymous survey first requested all parents to list the ages of their children aged 6 years or younger or indicate that they had no children of this age. Only parents with children 6 years or younger (the ages for which blood lead level screening is recommended in Illinois) were requested to complete the remainder of the survey. The next part of the survey instrument was the 24-question Chicago Lead Knowledge Test. The test included 5 questions related to general information about lead, 11 about lead exposure, 4 about prevention practices, and 4 about nutrition (Table 1). Additional survey questions concerned parental and child demographic data; housing conditions (rental status, age of home, ZIP code, number of years or months living at that residence, renovation or remodeling in past 6 months); whether respondents' children ever had an elevated blood lead level (defined as blood lead level ≥0.48 µmol/L [≥10µg/dL]); number of years attending that practice; and whether the respondent remembered receiving information about lead from a health care provider.
Subjects for this survey were parents of children presenting for health care in the summer of 1997 at any participating primary care practice (23 pediatric and 1 family practice) located in Chicago and its suburbs. Nine participating practices were in the city, including 4 health centers that serve mostly low-income minority clients. Twenty-two of 40 Pediatric Practice Research Group2practices invited to participate did so. Two practices that were not members of the Pediatric Practice Research Group also participated after physicians reviewed a draft of the Chicago Lead Knowledge Test and expressed an interested in evaluating parental knowledge about lead at their practice site.
Surveys were sequentially distributed to parents by office staff or a research assistant and were self-administered. When a study research assistant was at the office, if a parent requested, the research assistant obtained a response by private interview. Response rate was not determined. After data collection was completed, practices posted an annotated answer key.
Data were analyzed using SPSS for Windows, release 6.1.2 and SAS, version 6 (SAS Institute, Cary, NC). A Chicago Lead Knowledge Test score (the number of correct responses) was computed for each phase 2 respondent. Test questions skipped were considered to be a don't know response. The relationship between home age and receipt of lead information from a health care provider was examined using the Mantel-Haenszel test for linear association. Using 1-way analysis of variance (ANOVA), the relationship between individual variables (race, respondent's education, home ownership, age of home, ever had child with high blood lead level, recalled lead education from health care provider, insurance status, recent remodeling or renovation, and parental age) and the score was determined. In these analyses, response groups as outlined in Table 2were used. Age groups were selected so that approximately one third of respondents were in each group. The Scheffé test was used to determine similar groups within each variable; if similar groups were found and combining the groups seemed logical (for example, sequential age groups were combined, but differing race/ethnicity groups were not), groups were combined in further analyses.
To determine the relative importance of variables, a technique for developing an ANOVA model was applied. Using SAS, variables were entered into an ANOVA analysis in a forward stepwise fashion. Variables that did not significantly add to the model so developed were eliminated. All 2-way interactions were examined. Because particular education efforts may need to be applied to those with the most limited knowledge, a logistic regression model, using forward stepwise entry of variables, was applied to examine variables that differentiated those with the lowest Chicago Lead Knowledge Test scores (0-8 correct) vs those scoring the highest (16-24 correct); respondents with test scores of 9-15 were excluded from this logistic regression analysis.
Questions were developed and reviewed with the aid of multiple lead experts. Based on their suggestions, some questions were added (eg, relation of lead to water boiling, need for lead for good nutrition), wording of others modified, and several were deleted.
Fifty-one parents completed the final Chicago Lead Knowledge Test questions twice. Retests were completed after a median of 3 days (range, 2-4 days). Percent agreement between test and retest responses to individual questions was 88% to 100% (88% for 1 question, 90% for 1, and ≥94% for the other 22 questions). The Pearson product moment correlation for test scores between the 2 occasions was 0.96.
There were 2225 parents of children 0 to 6 years old who completed the survey. Six suburban practices contributed more than 200 responses each (range, 214-332) and 18 practices contributed less than 100 responses each (range, 9-90). Practices enrolled parents for a range of 2 to 30 days.
Ninety percent of the respondents were mothers. The mean age was 32.9 years (SD, 5.9 years). Eighty percent were home owners. Other characteristics of respondents and their children are listed inTable 3. Thirty-six percent of the 329 respondents who lived in a home built before 1950 and 28% of the 324 respondents who lived in a home built between 1950 to 1959 reported remodeling or renovation of their home in the past 6 months. Fourteen percent of respondents listed a Chicago ZIP code for their residence. The median duration of families residing at their current address was 4.0 years; 19% had lived at their current home for 1 year or less.
Overall, 34% of respondents remembered receiving lead information from a health care provider. This varied by practice from a low of 10% to a high of 79%. Respondents living in older homes were more likely to recall receipt of information about lead than those in newer homes or those who did not recall the age of their home (built before 1950, 41%; built 1950-1959, 37%; built after 1960, 32%; do not know home age, 31%;P<.01 [Mantel-Haenzel test]).
The mean score on the Chicago Lead Knowledge Test was 12.2 (SD, 3.7) of the possible 24 points. Test questions and percentage of correct responses are given in Table 1. General questions about lead and those related to lead exposure were more often answered correctly than questions about prevention or diet.
In 1-way ANOVA analyses, Chicago Lead Knowledge Test score was significantly associated with all variables tested (Table 2). In these analyses, higher scores were associated with recall of receipt of information about lead from a health care provider, having a college degree or higher, being 30 years of age or older, owning a home, living in an older home, knowing whether one of their children ever had an elevated blood lead level, indicating child's race/ethnicity as white vs African American (other contrasts were not significant), having non-Medicaid or self-pay health insurance for their child, and knowing whether renovation or remodeling had occurred in the past 6 months. Ninety percent of respondents had sufficient descriptive data to be included in the computation of the final ANOVA model. Five variables contributed to the model (Table 2; F=41.5,P<.001). There were no significant 2-way interactions. Other (unmeasured) factors also influence knowledge score, as this model only explained 14% of the variation in test score (multipleR2=0.14). After adjusting for other variables in the model, respondents who remembered receiving lead information from their health care provider scored 1.8 points higher than those who had not received information or were unsure. Similarly, college graduates scored approximately 1.5 points higher than those with less education; those aged 30 years or older scored 0.6 points higher than younger respondents; Hispanic respondents scored approximately 1.1 points higher than white, 1.1 points higher than other, and 1.8 points higher than African American respondents; and those living in homes built before 1950 scored 0.5 points higher than those in homes built in 1950 or later and 1.6 points higher than those not knowing home age.
Fifteen percent (337/2225) and 19% (417/2225) of respondents were in the lowest-scoring (0-8 correct) and highest-scoring (16-24 correct) groups, respectively. Eighty-seven percent and 89% of these respondents, respectively, had complete demographic data and were included in a logistic regression. The same 5 variables that differentiated score in the ANOVA model were the only variables that significantly differentiated those scoring high from those scoring low. Those in the high-scoring group were more likely to recall receipt of information from a health care provider (adjusted odds ratio [OR]=6.2; 95% confidence interval [CI], 4.0-9.6), have a college degree (adjusted OR=4.9; 95% CI, 3.2-7.3), be Hispanic vs white (adjusted OR=4.1; 95% CI, 1.8-9.1) (other contrasts not significant), be 30 years of age or older (adjusted OR=2.1; 95% CI, 1.3-3.3), and live in a home built before 1950 vs not knowing home age (adjusted OR = 8.3; 95% CI, 3.0-23.0) or a home built 1950 or later vs not knowing home age (adjusted OR = 5.4; 95% CI, 2.1-13.8).
The Chicago Lead Knowledge Test had good test-retest reliability and was used in a variety of inner-city and suburban practices. This new self-administered tool will aid in the comprehensive evaluation of lead education programs.
Many respondents were able to correctly answer questions about lead exposure. However, questions related to lead poisoning prevention and the relationship of nutrition to lead poisoning were less widely known. Higher scores were found for respondents who recalled receipt of information about lead from a health care provider. This should reassure physicians that parents are listening to and retaining at least some of what they say. It is logical and reassuring that parents living in older homes knew more about lead, because these children are at the highest risk. Recent renovation or remodeling did not add to the ANOVA model. Because renovation or remodeling occurred more commonly in older homes, it is likely that the variable home age also accounts for knowledge gained surrounding home renovation or remodeling. The finding that Hispanic respondents scored particularly well was unexpected. We believe the higher scores for the Hispanic respondents were a reflection of intensive lead education efforts at the practices they attended. Most Hispanic respondents were enrolled at inner-city practices that serve primarily low-income families who live in areas of deteriorating older housing. These practices routinely include repeated lead screening and lead education among the services they provide.
Adjusted mean score differences for contrasts within variables in the ANOVA model were approximately 1 to 2 points. The magnitude of these differences is small and clinical significance is uncertain. However, these are independent effects. When taken together (eg, the application of lead education to families in older housing), a more substantial increase in lead knowledge might be expected. The importance of these variables is further strengthened because of their ability to differentiate the highest- from the lowest-scoring respondents.
Two published studies have assessed parental lead prevention knowledge via in-person or telephone interview.3,4 In both studies, similar to our self-administered survey findings, parents had limited knowledge about ways to prevent lead poisoning and the importance of good nutrition as a preventive measure.
One third of the parents we surveyed remembered receiving lead information from their health care provider. This is similar to findings of Mahon4 and compares favorably with results reported by Chaisson and Glotzer.5 In their study of parents interviewed when receiving public health department notification that their child had an elevated blood lead level, only 12% (16/139) remembered receiving information about lead prevention prior to the identification of the elevated level.5
Our study results suggest that the greatest parental knowledge deficit seems to be in the role of diet in prevention of lead poisoning.6,7 Other preventive actions were also poorly known. Test results are clear; parents know about the risk of lead, but they do not know much about actions they can take to reduce lead exposure and absorption. Educational programs need to increase their preventive focus.
The requirement mandated by the Residential Lead-Based Paint Hazard Reduction Act of 1992 to inform tenants of known lead-based paint in a rental unit was known by less than half of the respondents. This notification requirement needs further dissemination.
In some busy inner-city clinics, we only administered the Chicago Lead Knowledge Test when a research assistant was present to aid those with reading difficulty. Because of the costs associated with this method, fewer inner-city respondents are included than we had desired. The anonymous nature of our survey and time and money limitations made response rate calculation impossible.
The results of this survey do not reveal how often children are being tested for high blood lead levels, nor did it allow us to compare the extent of parental knowledge with the prevalence of high blood lead levels in a child. We do not know if parental knowledge leads to application of prevention behaviors or prevention of childhood lead poisoning. This is an important area for future study.
The Chicago Lead Knowledge Test is a potential tool to aid in the evaluation of lead education programs. In this sample, parents who remembered receiving information about lead poisoning and its prevention from health care providers scored higher on the Chicago Lead Knowledge Test than other parents. Most parents had very limited knowledge about ways to prevent lead poisoning. Further research is needed to determine effective ways to increase parental knowledge of lead prevention and to determine whether enhanced parental lead knowledge decreases the incidence of childhood lead poisoning.
Editor's Note: I wonder how well this evaluation tool would work with parents who had less than a high school education (only 5% of this sample).—Catherine D. DeAngelis, MD
Accepted for publication July 6, 1998.
This study was supported by the Student Research Program of the American Pediatric Society and the Society for Pediatric Research, Elk Grove Village, Ill (Mr Mehta).
Members of the Statistical Sciences and Epidemiology Program at Children's Memorial Medical Center (Jean Gagliardi, Raquel Sandoval, Susan LeBailly, PhD, and Patricia McGuire) and Satish Charo assisted with study conduct. Heather Hastings, MPH (Project Manager, Neighborhood-Based Childhood Lead Primary Prevention Project), provided advice on question development. Katherine Kaufer Christoffel, MD, MPH, provided editorial review, Edwin Chen, PhD, provided statistical and editorial review, and Shiaoting Chiu helped with statistical analysis.
A copy of the Chicago Lead Knowledge Test and annotated answer sheet is available by request.
Andrew Davis, MD, Anchor Coleman, Chicago, Ill; John Poncher, MD, Pat Bechtel, RN, Associated Pediatricians,* Valpiso, Ind; Bonnie Typlin, MD, Child and Adolescent Health Associates,* Chicago; Jaye Schreier, MD, Child Life Center,* Homewood, Ill; Bennett Kaye, MD, Children's Healthcare,* Chicago; Elizabeth Hawkes, MD, Claretian Medical Center,* Chicago; H. Garry Gardner, MD, DuPage Pediatrics,* Darien, Ill; Jim Olson, MD, Robert Clarick, MD,Erie Family Health Center,* Chicago; Steven Stabile, MD,Family Health Center of West Logan Square, Chicago; Lori Walsh, MD, Linda Walsh, Glenbrook Pediatrics,* Glenview, Ill; Carl Toren, MD, MPH, Infant Welfare Society,* Chicago; T. Randall Kinsella, MD, Lake Forest Pediatrics,* Lake Forest, Ill; Diane Fondriest, MD, Lake Shore Pediatrics,* Lake Forest; Stephen L. Brookstein, MD, Medical Pediatrics Ltd,* Arlington Heights, Ill; David Dobkin, MD, North Arlington Pediatrics,* Arlington Heights; Richard Burnstine, MD,North Suburban Pediatrics,* Evanston, Ill; Marc Weissbluth, MD, Rebecca Unger, MD, Northwestern Children's Practice,* Chicago; Daniel Lum, MD, Pediatric Associates of the North Shore,* Wilmette, Ill; Norman Segal, MD, Ramona Walker,Pediatric Healthcare Associates,* Buffalo Grove, Ill; Barry Altshuler, MD, Pediatric Specialists,* Barrington, Ill; Timothy Wall, MD, Pediatric Health Associates,* Naperville, Ill; Lena Sen, MD, Evanston; Edward Traisman, MD, Traisman, Benuck, Traisman, & Merens,* Evanston; Barbara Bayldon, MD,Winfield-Moody Health Center,* Chicago.
Members of the Pediatric Practice Research Group are marked with an asterisk.
Corresponding author: Helen J. Binns, MD, MPH, Children's Memorial Medical Center, 2300 Children's Plaza, 208, Chicago, IL 60614 (e-mail: firstname.lastname@example.org).
Samir Mehta, Helen J. Binns. What Do Parents Know About Lead Poisoning?The Chicago Lead Knowledge Test. Arch Pediatr Adolesc Med. 1998;152(12):1213–1218. doi:10.1001/archpedi.152.12.1213