VI. Universal versus Targeted Screening
Universal testing, especially when the testing is affordable and accessible, and when providers and parents are sufficiently engaged, can lead to high screening rates to identify more children with high BLLs. Testing according to targeted requirements and even recommendations, when they are strong and clear, reinforced by educational efforts, and supported by the medical community, can also lead to a successful screening program, although universal screening is still preferable.
CDC moved toward targeted testing as higher levels could be linked to specific locations in a state. By now emphasizing targeted testing, CDC encourages states to use common knowledge about lead sources and their own data to develop plans that test children who are most likely to be exposed to lead. However, that can be a risky approach when not all lead sources are known. The potential dangers of lead pipes were largely thought solved until recently. Lead in candy can also be a source, but some state questionnaires don’t ask if the child eats imported candy. Lead in soil from industry can contribute to EBLLs, but questionnaires don’t always ask whether children live near current or former lead-emitting industries. Targeted screening is only as good as the criteria chosen to determine the target.
The Utility of Questionnaires in Targeted Screening
Many states (39 + the District of Columbia) recommend or require that doctors administer risk assessment questionnaires or at a minimum ask questions about risk to determine whether to test children for lead. Examples of the kinds of questions commonly asked include:
- Does the child live in or frequently visit a building built before 1950?
- Does the child live in or frequently visit a building built before 1978 that was recently renovated?
- Does the child’s sibling or playmate have lead poisoning?
- Is the child an immigrant, refugee, or foreign adoptee?
- Does the parent or caregiver have a job or hobby involving lead?
- Does a member of the household use ethnic remedies, cosmetics, imported pottery with a lead glaze, or imported candy?
These questions cover many sources of lead – lead paint in older housing, especially if the paint is deteriorating and being disturbed during remodeling; lead brought into the house by an adult who may have lead on their shoes or clothing; and lead from other countries, where the regulations lag behind the U.S.
But not all states actually use all of these types of questions in their risk assessments. California, for example, only requires testing for children receiving services from publicly supported programs such as California’s Medicaid program and WIC, and for those spending a lot of time in a structure built before 1978 that was recently renovated or where the paint is in disrepair. Doctors in California aren’t required to ask if children have eaten candy manufactured outside the U.S., when this could be a serious source of lead poisoning. Staff in the Nevada Department of Health and Human Services noted a concern about lead in candy imported from Mexico. Most universal screening states found some lead exposure came from imported items, such as food coloring, spices, makeup, or pottery.
Even states with more extensive questionnaires may not cover sources that could be important. Only very few address lead in water from lead pipes or plumbing fixtures. Lead pipes weren’t banned from use in new installations or repairs until 1986, and even then, pipes were still allowed to have 8% lead until 2014. While water is not generally the largest source of lead exposure, it can add to other sources of lead in the environment and raise BLLs enough to cause serious health effects, as has happened in Flint, Michigan.
While it is generally acknowledged that questionnaires can be an effective tool for finding BLLs above 10 µg/dL, CDC’s previously designated “level of concern,” they aren’t always effective, as the CDC’s Advisory Committee on Childhood Lead Poisoning Prevention concluded after analyzing several studies. One reason for the ineffectiveness may be that all parents don’t know the age of their homes, which is a standard question used to determine risk in lead risk assessments. A 2003 study compared parents’ responses to “Does your child live in or regularly visit a house that was built before 1950?” with the age of their house, ascertained via tax assessor records. Overall, just over half of the parents who should have responded yes based on their house’s age did in fact “accurately [report] this exposure.” Fewer than half of the parents of children with Medicaid insurance who lived in pre-1950s housing answered correctly.
Further, risk assessments may not be effective in identifying children with BLLs under 10 µg/dL. A 2012 study published in Public Health Nursing found that verbal lead risk assessment questionnaires were not effective in consistently detecting measurable BLLs in children. In fact, the group of children whose parents answered “no” to all questions and would not normally be tested, had slightly higher, although not significantly different, average BLLs than those with at least one “yes” or “don’t know” answer (2.2 µg/dL v. 1.6 µg/dL). Although the higher average BLL was low, studies have found health impacts even at 2 µg/dL. The overall finding of the Public Health Nursing study was that universal testing is necessary to detect measurable BLLs so the sources can be found and eliminated.
Another concern with relying on questionnaires is that it’s difficult to ensure doctors administer them. Most states only recommend risk assessments, but even for the states that require them, it may be difficult to monitor compliance aside from when it triggers a BLL test that is subsequently reported.
Maryland’s Decision to Move Away from Targeted Screening
Testing children in pre-determined high-risk zip codes is another popular approach to targeted screening. Until 2016, Maryland required testing (without a risk assessment first) only for children living in a limited number of at-risk zip codes.
Maryland amended its regulations to require universal testing for 12- and 24-month-old children born in 2015 or after, by incorporating by reference the “Maryland Targeting Plan for Areas at Risk for Childhood Lead Poisoning,” issued in October 2015 by the state Department of Health and Mental Hygiene (DHMH). This plan designates the entire state as “at risk” for lead exposure for three years. The state plans to reevaluate whether universal testing is still necessary once three years of data have been collected.
In its 2015 Targeting Plan, DHMH discussed how progress has been made toward the state’s goal of ending lead poisoning, but children still face exposure to lead, which “can cause permanent neurological damage that may be associated with learning disabilities, decreased intelligence, and behavioral problems.” The agency indicated that blood lead testing is an important part of an overall lead poisoning prevention strategy, because interventions can be provided once children with lead exposure are identified.
The agency listed key reasons for changing the testing program from targeted to universal, including that the risk factors for elevated levels had changed and now included non-paint sources and sources that could not be identified. Additionally, with the CDC’s recommendation of the new, lower reference level in 2012, DHMH surmised that more children could be at risk and should be tested. The agency wanted a testing strategy that was more effective in identifying children at risk to complement and strengthen the state’s continuing primary prevention efforts.
To determine the best type of screening plan, DHMH considered universal testing and two different targeted screening plans – one based on the zip code distribution of past BLL test records, and the other on the geographical distribution of traditional risk factors such as housing age and poverty level. Compared with universal testing, the department estimated that both targeted models would miss a significant number of children expected to have BLLs of 5 µg/dL or higher.
In other words, non-universal testing strategies exclude some at-risk children and don’t create a representative picture of lead levels. Using Maryland’s historical data to create a testing plan could bias testing toward previously defined at-risk areas, since fewer children outside those areas were tested. Plans based largely on housing and demographics minimize the role of non-housing sources of lead (and by inference, they also ignore the fact that some sources of lead exposure cannot be identified). DHMH also noted that having universal testing for a limited time period “is easier and simpler to implement and communicate, and will provide useful data on the true prevalence and distribution of children with elevated blood lead levels in the State.”
While universal testing and associated follow-up is expensive – DHMH estimated they ranged from around $4 million to $6 million for Maryland – the costs of lead poisoning from not finding all children with elevated levels and addressing the sources are high. Beyond the health impacts, lead poisoning imposes economic costs, including those related to “lifetime earnings, tax revenue, special education, [and] criminal justice” because of lead’s impact on IQ and on physical health. According to DHMH, reducing all BLLs of 5 µg/dL or higher in those aged one and two in the state, identified through testing, would lead to cost savings of between $143 and $556 million. DHMH translated this into a return of $24-142 for each dollar invested. Addressing elevated lead levels has significant societal benefits, both for children’s health and wellbeing, and for society as a whole.
Lessons for Other States
Targeted screening has limitations. Most states use questionnaires to help doctors determine who to test, but their thoroughness varies and even the more rigorous ones may not identify all children with detectable or even elevated levels of lead. It is also difficult to ensure that they are used. Since Maryland has adopted universal testing for children born in 2015 and after for three years, testing decisions at 12 and 24 months for this population are not based on a risk assessment. However, the assessment is still required at well-child visits for other purposes.
Maryland looked closely at two targeted screening models – based on historical BLL tests or demographics – and determined neither would suffice. It’s likely that other states are in the same situation as Maryland, with significant numbers of children with elevated lead levels potentially overlooked by their current non-universal testing plan. In the next three years, Maryland will be able to establish a much more complete baseline of data and could have a strong foundation to move back to targeted testing if they find patterns to justify it.
Universal testing is also easier for doctors to follow – in Maryland’s case, doctors are required to test all children born in 2015 and after at 12 and 24 months. They don’t have to remember which zip codes they’re supposed to test in.
States may have concerns about costs, and it is important to account for this in the development of a lead screening policy. Universal testing and associated follow-up is more costly than targeted testing, as indicated above. However, the long-term benefits, both financial and health-protective, clearly outweigh the short-term costs.