Methods

The interactive 40 minute MORE HEALTH teaching session included prelesson materials for teachers—an “ALWAYS ASK FIRST” video for the kindergarten lesson, and a vocabulary list and books for the third grade lesson. For the kindergarten lesson, the instructor used singing, puppet shows, a story board, and poison lookalikes. For the third grade lesson, the instructor supplemented didactic instruction with class discussion, poison lookalikes, an experiment to demonstrate “invisible” gases, and opportunities for the children to act out scenarios of contacting the poison center. The students also discussed a poison proofing checklist and their responsibility to help poison-proof their homes. To reinforce what was discussed in class, kindergarten children were provided with materials to take home that included a poison center telephone sticker, a poison patrol badge, a letter to parents, and a brochure. Similar materials, including a magnet with the phone number of the poison center, were given to the third grade students. Also, posters of poisonous plants and snakes were distributed to participating schools.

The evaluation was based on a post-test only control group design. Three schools were selected to serve as evaluation sites and three others as control sites. The School Board required that their office and the MORE HEALTH Program choose the intervention and control schools. The agreed upon criteria for selection included schools that were representative of all socioeconomic levels and whose schedules permitted the researchers access to the classrooms.

Before the evaluation, two researchers reviewed videotapes of the lessons and attended actual presentations to confirm the instructor's fidelity to the lesson plan. All presentations were given by the same instructor. To assess the efficacy of the lessons, age appropriate questionnaires were developed based on a review of the literature. Questions included both forced choice and open ended items. Content validity was established by health educators experienced in survey design and poison prevention. Items were read aloud.

Interviews were administered one-on-one during times that created the least disruption. The same researcher conducted all interviews. She was introduced by the teacher and built rapport before asking questions, to which the students responded orally. Each interview took approximately 10 minutes to complete. No students in the intervention or control groups had previous school based lessons that focused on poison prevention.

Interview topics included: (1) poison introduction (for example, “What is a poison?” and “Where are poisons found?”); (2) naming poisonous spiders and snakes and knowing what to do if bitten; (3) identification of poison lookalikes and whether to eat or drink lookalikes; (4) performance of the stingray shuffle (that is, students shuffle their feet to simulate what they would do in water to avoid being stung by a stingray); (5) knowing what the poison center telephone sticker is and where to place it in the home; (6) poison vocabulary (for third grade only: examples include knowing the forms of poisons and how poisons enter the body); and (7) how to make the home safe for younger brothers, sisters, or friends (third grade only).

The interview questions were pilot tested with kindergarten and third grade students at a school that was not part of the study to assess feasibility, timing, comprehension, and logistical issues. Post-test interviews were administered between one and two weeks after the intervention. A questionnaire was also developed for the parents of intervention children. This was included with the children's post-lesson packet, with an accompanying letter asking parents to return the questionnaire to the teacher. Parents were asked if they took steps to make sure their homes were safe from various poisons, and if they were going to take additional steps after reviewing their child's take home materials. Questions focused on proper storage and disposal of harmful products, identification of poisonous plants, possession of syrup of ipecac, having the poison center emergency sticker on or near the telephone, description of any poisoning incidents in the home reported to the poison center, and perception of the quality of the center's response.

All children and parents were informed that their participation was voluntary and that confidentiality would be maintained. No children refused to be interviewed. The project was approved by the University's Review Board and the School System. Data analysis included frequencies, descriptive statistics, and calculations of the odds ratio and 95% confidence interval (CI) to determine the odds of answering questions correctly based on being in an intervention or control group. Exact upper and lower limits were calculated in instances where Cornfield 95% confidence limits were not accurate.

Results

One hundred ninety four intervention children and 184 control children were studied: 84 intervention and 71 control kindergarten children and 110 intervention and 113 control third grade children. Intervention children consistently answered more questions correctly than children who were part of the control groups (tables 1 and 2).

To determine the probability of children answering questions correctly in the two groups, we computed the odds ratio and corresponding 95% CI. The differences for kindergarten children were statistically significant for all items for which odds ratios could be computed except one (naming poisonous snakes). For the third grade children, differences were statistically significant for all but two items—picking up strange snakes or spiders and not taking medicine without an adult present.

Seventy seven intervention school parents (39.7%) returned the questionnaire. Table 3 shows selected responses of kindergarten and third grade parents. The majority of parents reported that their homes were poison-proof and the only item where fewer than 50% of the parents responded correctly was having syrup of ipecac available. Ten parents reported having to contact the poison center about a poisoning incident and all reported that the center responded to their satisfaction.

Discussion

To our knowledge, this study is one of the very few examples of an evaluation of a poison prevention lesson aimed at children in the primary grades and results show that this intervention changed knowledge significantly for both grade levels studied. Key concepts, such as the identification and placement of poison center stickers with emergency telephone numbers, can be communicated effectively to young children. This may be particularly important to the success of community-wide education campaigns because schools can serve as important conduits for prevention messages.

Though the results generally support the intervention, not all concepts were absorbed uniformly. For example, in the kindergarten group, children did not retain information about poisonous spiders. In addition, some questions may have been “too easy” to test true gains in knowledge. All children in kindergarten showed knowledge about avoiding contact with “strange” snakes and spiders, and not taking medicines without adult supervision.

A similar situation applies to the third grade results. All of the intervention group correctly identified the proper protocol to follow if bitten by a snake or spider, as did 76% of the control group. All intervention subjects knew what to do if a sibling was poisoned, as did nearly 78% of controls. The third grade program's major shortcoming appeared to be its inability to communicate the various forms in which poisons exist (for example, gas, liquid, solid, etc). However, at post-test, more than 25% of the intervention subjects answered the item correctly, compared with none of the controls.

Study limitations

Because of the requirements of the school system, it was not possible to randomize the schools. However, the post-test only control group design eliminates the effects of pretesting as a confounding factor in interpreting results,²¹ and decreases the labor intensity of the evaluation effort.²² Although complete equivalence of the intervention and control groups before the intervention cannot be guaranteed with this design, the assumption of equivalence is not an unreasonable one, given the age of the children and the documented absence of formal school based education efforts. It was not possible to do second interviews to determine if knowledge was retained. Most importantly, we were unable to study the children's behavior to see if their knowledge translated into action. However, an unpublished study from Georgia suggests that a similar curriculum significantly changed children's knowledge, attitudes, and practices regarding poison safety (Paige Cucchi, Georgia Poison Center, personal communication, July 1997). In addition, it has been shown that increases in knowledge significantly contribute to intention to use poison prevention behaviors.²⁰

The results of our study may not be generalizable to other groups and settings. However, they are indicative of the potential for early intervention programs to stimulate awareness among children and their parents.

For school based poison prevention programs to improve and expand, it is recommended that lessons be incorporated into other subject areas and that parent communication occurs. In addition, it is important that schools partner with community injury prevention groups and poison centers. Programs developed and implemented need to be evaluated for accountability, improvement, and decision making.²³ There is a need for future research studies to determine, through randomized controlled trials, if and how education for young children significantly contributes to behavior change and decreases in poison related morbidity and mortality.