Lead Screening Practices of Pediatric Residents

Objectives  As part of their training, pediatric residents provide primary care services to young children, including youngsters who may have elevated blood lead levels. We set out to (1) determine the percentage of pediatric residents who screen children for elevated blood lead levels according to the guidelines of the Centers for Disease Control and Prevention and the American Academy of Pediatrics; (2) assess the likelihood of lead screening by residents based on demographic and practice-setting characteristics; and (3) compare the attitudes of residents who report that they are universal screeners, selective screeners, or nonscreeners.

Design  Confidential, cross-sectional survey of a nationally representative sample of pediatric residents conducted as part of the American Academy of Pediatrics 28th Periodic Survey of Fellows.

Subjects  One hundred forty-three responding pediatric residents (51% response rate).

Results  Seventy-five percent of pediatric residents reported screening all patients aged 9 to 36 months for elevated blood lead levels, 21% reported screening some, and 4% reported screening none. Pediatric residents who cared for patients in urban settings were more likely to report screening patients for elevated blood lead levels than were pediatric residents who cared for patients in suburban or rural settings (100% vs 73%; P<.001), and pediatric residents in the Northeast were more likely to report screening universally than were residents in the rest of the country (93% vs 63%; P<.001). Overall, pediatric residents who reported screening patients universally were more likely to believe that the benefits of screening outweigh the costs than were residents who reported screening patients selectively (67% vs 17%; P<.001).

Conclusions  Most pediatric residents reported that they screened patients for elevated blood lead levels, either universally or selectively. Nevertheless, the screening practices of pediatric residents and their opinions concerning the relative benefits and costs of lead screening largely reflect the areas of the country and the practice settings in which they had their primary care experiences.

IN 1991, the Centers for Disease Control (CDC), Atlanta, Ga, revised its lead screening guidelines and called for frequent assessment of the risk of lead exposure and universal lead screening for children younger than 6 years.¹ In 1993, the American Academy of Pediatrics (AAP), Elk Grove Village, Ill, adopted similar guidelines.² A recent nationwide survey³ indicated that approximately half of US pediatricians report screening all patients younger than 6 years for elevated blood lead levels, and more than 90% report screening at least some patients. However, in the same survey, many pediatricians also expressed reservations about the benefits of screening to identify children with elevated blood lead levels.

Residency offers an opportunity for physicians-in-training to become aware of and implement appropriate screening practices as part of their primary care experiences. This study was undertaken to determine the extent to which pediatric residents screen patients for elevated blood lead levels and evaluate children for possible lead exposure.

The AAP 28th Periodic Survey of Fellows, conducted in late 1994 and early 1995, evaluated pediatrician practices and opinions regarding lead screening and was sent to a random sample of 1610 AAP members. Of the 1035 respondents, 734 provided primary care, including 143 who were pediatric residents at the time of the survey. This study focuses on the survey responses of these 143 residents.

The survey instrument was an 8-page, self-administered, forced-choice questionnaire. The questionnaire was made up of 3 sections: (1) demographic and practice-setting characteristics; (2) screening and educational practices; and (3) opinions regarding issues related to elevated blood lead levels. Opinion questions used a 5-point Likert rating scale (strongly agree, agree, neither agree nor disagree, disagree, strongly disagree).⁴

Bivariate analyses were used to summarize and compare responses of residents by demographic and practice-setting characteristics and to compare the proportions of respondents categorized as universal screeners, selective screeners, and nonscreeners who agreed with each of the opinions regarding blood lead screening. χ² Analysis was used to test the significance of categorical variables. The Mann-Whitney test was used to compare the estimated proportions of children with elevated blood lead levels in the practice settings of universal screeners, selective screeners, and nonscreeners.

Multivariate logistic regression analysis was used to identify independent demographic and practice-setting characteristics that differentiate pediatric residents who reported screening patients universally and pediatric residents who reported screening patients selectively or not at all. The independent variables for the regression model were selected by conducting bivariate comparisons between the aforementioned dichotomous variable and the following characteristics: gender, geographic area, location of the resident's primary care practice, and year of training; only comparisons with P<.10 in the bivariate analyses were included in the regression model. Logistic regression analysis was used to calculate odds ratios and 95% confidence intervals for characteristics associated with whether or not pediatric residents reported screening universally.

The 143 pediatric residents who completed the survey represent 51% of the 281 residents surveyed. Respondents were distributed demographically as shown in Table 1. One hundred seven (75%) reported screening all patients younger than 6 years for elevated blood lead levels, 30 (21%) reported screening some patients, and 6 (4%) reported screening none. Reported screening practices varied by geographic area and by the type of community in which the residents had their practices (Table 2). Pediatric residents in the Northeast were more likely to report screening universally than were those in the remainder of the country (93% vs 63%; P<.001). In addition, pediatric residents who cared for patients in urban settings were more likely to report screening all or some patients for elevated blood lead levels than were pediatric residents who cared for patients in suburban or rural settings (100% vs 73%; P<.001).

Backwards stepwise logistic regression indicated that doing a residency in the northeastern United States (odds ratio, 3.00; 95% confidence interval, 1.58-5.67) and being in the third or fourth years of residency training (odds ratio, 1.79; 95% confidence interval, 1.08-3.01) independently predicted whether pediatric residents reported screening universally. In combination, area of the country and year of residency training correctly predicted whether responding pediatric residents reported that they were universal screeners or nonuniversal screeners in 76% of cases.

Pediatric residents who reported screening universally for elevated blood lead levels estimated that a mean proportion of 14% of the children younger than 6 years in their practice settings had elevated levels (ie, ≥0.48 µmol/L [≥10.0 µg/dL]) compared with 10% of children cared for by residents who reported screening selectively and 3% of children cared for by residents who reported not screening (universal screeners vs nonscreeners, P=.04; universal screeners vs selective screeners, P=.25; selective screeners vs nonscreeners, P=.10). Among the residents who reported screening patients selectively (n=30), the most common factors prompting them to screen particular children were a sibling with an elevated blood lead level (84%), a history of pica (81%), a child who lives in an older home with peeling paint (78%), a parental request for screening (76%), a history of toxic substance ingestion (71%), and information that a child's parent uses lead in a job or hobby (71%). Although these prompts for screening largely mirror items on the CDC lead exposure risk assessment questionnaire, only approximately half of the residents who reported screening patients selectively used a questionnaire to consistently evaluate risk in a comprehensive manner.

Among pediatric residents who reported universally or selectively screening patients for elevated blood lead levels, 68% reported screening most children 9 to 12 months of age and 73% reported screening most children 13 to 24 months of age. The median age at which these residents reported that they first screened high-risk patients was 9 months, and the median age at which they reported that they first screened low-risk patients was 12 months. All of these residents reported screening children at well-child care visits if they believed that the children needed testing, and 13% also reported screening at acute care visits, 46% at follow-up visits for acute illnesses, and 63% at follow-up visits for chronic illnesses.

Of residents who reported screening patients for elevated blood lead levels, 88% also reported providing information to parents about the dangers of lead and ways to prevent lead exposure. Sixty-one percent responded that they provide this information before screening is performed, and 39% stated that they only provide this information to parents if the child has an elevated blood lead level. By report, information is usually given concerning potential sources of lead (99%), the detrimental effects of lead (95%), and methods to limit lead exposure (93%).

Table 3 shows the opinions of residents regarding issues related to blood lead screening. Most pediatric residents believe that parents may become anxious if their children have blood lead levels of 0.48 to 0.92 µmol/L (10-19 µg/dL). Residents also agreed that epidemiological studies should be performed to determine which communities have significant proportions of children with elevated blood lead levels. However, most residents (54%) also believe that the benefits of lead screening outweigh the costs.

Table 4 shows the proportions of residents characterized as universal screeners and selective screeners who agree with each of the positions listed in Table 3. Those who reported that they were universal screeners were significantly more likely than those who reported that they were selective screeners to believe that the benefits of lead screening outweigh the costs (67% vs 17%, P<.001). Pediatric residents in the Northeast were also significantly more likely than residents in other areas of the country to agree with that statement (65% vs 47%, P=.04), as were residents who reported that elevated blood lead levels were common among their patients (69% vs 47%, P=.03). However, year of residency training and the location of the resident's primary care practice (urban vs suburban or rural) were not associated with positions regarding this belief.

The results of this national survey of pediatric residents indicate that, at the time of the survey (1994-1995), most pediatric residents seemed to be screening young children for elevated blood lead levels in compliance with the guidelines of the CDC and the AAP.^1,2 Indeed, most pediatric residents responding to this survey reported screening all patients aged 9 to 36 months for elevated blood lead levels, while smaller proportions reported screening selectively or not at all. Universal screening was most likely to be done by residents whose practices were located in urban areas or in the northeastern United States, where the highest percentages of children with elevated blood lead levels are located.⁵ Moreover, most residents who reported that they screened selectively based their screening decisions for individual children on risk factors similar to those recommended by the CDC¹ (although a formal risk assessment instrument was used by only about half of these residents).

Overall, it seems that the views of residents regarding lead poisoning and lead screening issues generally coincide with most aspects of the CDC and AAP guidelines. More than 70% of responding residents agree with the CDC's definition of an elevated blood lead level as being 0.48 µmol/L or higher (≥10 µg/dL),¹ and 62% believe that interventions are effective in children with blood lead levels of 0.48 to 0.92 µmol/L (10-19 µg/dL). However, 84% believe that parents may become anxious if their children have blood lead levels in this range, and 91% believe that more epidemiological studies should be performed to identify communities that may have considerable proportions of children with elevated blood lead levels. It is clear from these latter responses that residents, like other practicing pediatricians,^3,6 have concerns about the usefulness of universal lead screening. Presumably, epidemiological studies would identify communities in which few children have elevated blood lead levels and in which universal screening is unlikely to be useful or cost effective. Indeed, based on studies showing variability in the prevalence of elevated blood lead levels in different geographic areas and among different population groups,^7-11 the CDC has recently proposed that blood lead screening be focused on children who are at the highest risk of lead exposure.¹² As a result, in most communities, universal blood lead screening will be replaced by selective screening.

In the future, children who are at high risk for lead exposure will continue to need close supervision and frequent assessment of blood lead levels. Just as providers often miss opportunities to immunize undervaccinated children,^13,14 it is easy to miss opportunities to obtain blood lead assay samples from high-risk children who are overdue for testing. Many such children visit providers for acute illnesses, at which time a blood lead assay sample may conveniently be drawn or ordered. However, this study and others^14,15 have shown that only a small proportion of physicians screen for elevated blood lead levels at acute care visits, thereby missing opportunities to perform blood lead level testing on otherwise hard-to-reach children.

The issue of parental anxiety prompted by having a child with a mildly elevated blood lead level has been raised previously by other physicians.^16-18 Although some parents may consider such a child to be vulnerable or may worry excessively about potential long-term effects of lead exposure, other parents may respond by taking measures to prevent further lead exposure.

Although most pediatric residents responding to the survey believe that interventions are effective for children with blood lead levels of 0.48 to 0.92 µmol/L (10-19 µg/dL), there is widespread concern in the larger pediatric community that there may be little to offer parents of children with mildly elevated blood lead levels.^7,8 To date, the results of most intervention strategies^19,20 for children having mildly elevated blood lead levels have been disappointing, although the results of at least 1 study suggest that interventions may be effective.²¹

Overall, despite predominantly caring for children who live in urban settings who may be at high risk, only a small majority (54%) of residents reported that they believe that the benefits of lead screening outweigh the costs. However, a significantly larger proportion of pediatric residents who reported screening universally agree with this position than do pediatric residents who reported screening selectively or not at all. It is impossible to discern from the survey whether opinions about the relative benefit and cost of lead screening predominantly reflect personal experience, the influence of supervising attending physicians, or other considerations. The fact that pediatric residents training in the Northeast were significantly more likely to believe that the benefits of screening exceed the costs implies that residents' positions regarding this question may reflect the frequency with which they care for children with elevated blood lead levels. Also, it seems that as residents progress in their residency training, they become more likely to universally screen patients for lead. Personal experience accumulated over time and increasingly ingrained practice patterns probably account for this phenomenon.

Certain limitations may have affected this survey. Relatively small numbers of residents were surveyed, and few respondents reported that they were nonscreeners, making it difficult, in particular, to characterize the nonscreener group. In addition, it is possible that nonrespondents had practices and opinions regarding lead screening that were substantially different than those of respondents. Finally, the practices and opinions noted in the survey do not directly reveal the attitudes of the residents who completed the survey, but rather only imply what their attitudes may have been. Nevertheless, it is likely that their responses to the survey items reflect the practices and opinions of a representative group of the next generation of pediatricians.

In 1994 and 1995, high proportions of residents screened young children for elevated blood lead levels in accordance with CDC and AAP guidelines. This represents a substantial change from the results of a 1989 survey of residency directors, which showed that, at that time, pediatric residents in only a small percentage of training programs screened patients for elevated blood lead levels in accordance with the guidelines then in effect.²² Universal screening recommendations and the widespread attention given to the 1991 CDC lead screening guidelines probably account for this change.

It is likely that most practice settings where residents care for patients have policies about the ordering and timing of screening tests, including blood lead screens. Residents would be expected to follow the lead screening policies of their practice settings or the screening recommendations of the attending physicians who supervise them. Indeed, the guidance of supervising physicians and their own patient care experiences may be more important than other factors in determining residents' opinions and practices. This is particularly significant because residents are likely to take the opinions and practices formed during their residency training with them when they complete that training and enter pediatric practice.²³

Accepted for publication October 7, 1997.

We thank Karen G. O'Connor and Gretchen Fleming of the AAP, J. Routt Reigart of the AAP Committee on Environmental Health, and Peter Briss of the Lead Poisoning Prevention Branch at the CDC for their assistance with the development and implementation of the survey.

Editor's Note: One of the reviewers for this study referred to lead toxicity as a disappearing problem. I hope that he is correct, but I fear we might stop appropriate screening to achieve the illusion of disappearance.—Catherine D. DeAngelis, MD

Corresponding author: Stanley J. Schaffer, MD, Department of Pediatrics, University of Rochester Medical Center, PO Box 777, 601 Elmwood Ave, Rochester, NY 14642.