Katrina Smith Korfmacher, PhD

Outreach Coordinator
Environmental Health Sciences Center

University of Rochester

September 2, 2003

 

Analysis of dust wipe sampling in 1998-2001 Monroe County HUD grant

 

Introduction

 

In 1998, the Monroe County Health Department was awarded a 1.7$ million Lead Hazard Reduction Grant.  The grant supported lead hazard reduction in 72 housing units in the City of Rochester over a period of three years.  Units that housed a lead poisoned child (blood lead level above 20 µg/dL) were prioritized for inclusion in this project (“EBL-units”).  Many of these “EBL units” were in multi-family houses, in which case the other units in the house (“non-EBL units”) were also included in the program.  The non-EBL units were included based on the assumption that the multiple units had similar paint and maintenance histories, in which case the non-EBL units were also likely to pose a lead risk to children.

Before the HUD-funded work was initiated, the EBL units had previously been investigated for deteriorated lead paint, repaired by the property owner, and issued a lead hazard reduction letter by the Monroe County Health Department’s Childhood Lead Poisoning Prevention Program (Lead Program).  The health department’s investigations are guided by Part 67 of the New York State sanitary code.  The health department issues a lead hazard reduction letter after visually confirming that the required repairs have been made.  In accordance with Part 67, the health department does not measure dust lead levels as part of this EBL investigation process.

As context for this report, it is important to understand the differences between the state’s program for reducing lead hazards in children’s homes under Part 67 (“state program”) and the US Department of Housing and Urban Development (HUD) lead guidelines and standards.  The state program’s focus is on protecting children who have been poisoned by lead paint.  When a child has a blood lead level over 20 µg/dL, the county health department is required to investigate the child’s home environment, which may include one or more housing units (including relatives, babysitters, etc.).  The investigator looks for the sources of lead poisoning, primarily lead-based paint in areas accessible to the child.  The regulatory process involves identification of deteriorated leaded surfaces and a Notice and Demand to abate the conditions conducive to lead poisoning.  The remediation is visually evaluated to determine whether or not the violations have been addressed. 

The HUD approach to lead is focused on making a home lead safe, as defined as having no deteriorated lead paint, no bare soil with lead levels over 400 parts per million, and no dust lead levels that exceed the EPA standards (40 µg/ft2 on floors, 250 µg/ft2 on window sills, and 400 µg/ft2 on window wells).  The HUD approach begins with a risk assessment that tests deteriorated paint surfaces, soil in bare areas of the yard, and levels of lead in dust in the house (between 8 and 12 samples are usually taken).  As in the state program, deteriorated lead based paint in areas accessible to children is addressed using either abatement or interim controls.  However, in the HUD approach, workers must be trained in lead safe work practices, and bare soil must be treated.  A “clearance” based on visual inspection and another set of dust wipe tests indicates whether or not the unit is lead safe at the time of testing.  HUD guidelines also recommend ongoing maintenance and monitoring of interim controls by the property owner and annual reevaluation of dust lead levels to insure the continued lead safety of the unit.

These and other differences mean that it is theoretically possible for a state program investigation to yield a unit that may not be safe by HUD standards.  There are three reasons a treated unit (under either program) may not pass a clearance test.  First, all sources of lead contamination may not have been addressed (for example, external sources of dust).  Second, work may have been done improperly – it is essential to follow lead safe work practices while doing and cleaning up after work that disturbs lead paint in order to avoid creating or making lead dust problems worse.  Third, new leaded dust may have been tracked into the unit from outside after the work was completed but before dust wipes were taken.  Because of the additional elements in the HUD approach (such as addressing soil and testing residual lead dust levels), the HUD procedures are more protective of children, although a HUD clearance test does not guarantee future lead safety as interim controls deteriorate or lead dust is tracked in from outside

As a result of the HUD-grant recruitment process that included non-EBL units in the same building as units that housed an EBL child, this HUD grant program provides a unique paired sample data set incorporating both the state and HUD frameworks for addressing lead hazards in homes.  Approximately half of the housing units in the sample had recently been investigated by the Monroe County Health Department using the state framework; the other half had not.  Although lead levels in dust were not measured as part of the EBL investigation, dust lead levels were measured as part of the HUD grant program assessment process.  Thus, it is possible to compare the dust lead levels in units the Health Department had previously investigated (EBL units) with those that had not been investigated (non-EBL units) prior to any HUD grant intervention. 

While this data set provides a unique opportunity to analyze dust lead levels in units after county health department investigations guided by the state approach, it is not a rigorous evaluation of the efficacy of the health department’s work.  Most significantly, the design is quasi-experimental in that it uses an existing natural control (non-EBL units) for the treated units (EBL-units).  However, we cannot be certain that there were not systematic differences between the EBL and non-EBL units that could confound results.  An experimental design that rigorously controlled for all differences between units would yield more certain results.  In addition, the data were not collected and controlled for the purpose of this analysis.  As further discussed below, this introduces additional uncertainties. 

Nonetheless, analyses of natural experiments such as this can yield useful suggestions for policy purposes.  For a variety of ethical, legal, and logistical reasons, it would be difficult to carry out experimental research on dust lead levels after EBL investigations.   Thus, while this analysis is not based on rigorously controlled experimental research, it yields insights that may help inform future decisions about how best to protect children who have been lead poisoned from living in homes with potential lead hazards.

 

Methodology

             

Dust wipe samples were reported by the county HUD program as averages by surface.  That is, multiple floor samples (usually between four and six) were taken in the unit and the average reading reported as the unit’s “floor dust lead level.”   The same process was used for window sills and wells.  This means that individual samples may have been higher or lower than the average reading; however, the average gives an overall indication of the lead hazards in the unit.  These average dust levels were analyzed separately for floors, window sills, and window wells.

The data were separated into EBL and non-EBL units.  Only units in which a lead hazard reduction letter had been issued within the past 12 months were included in the EBL sample; units that had been investigated more than 12 months before dust testing were included in the non-EBL sample.  There were 31 EBL units and 35 non-EBL units in the sample.  The average number of months between lead hazard reduction letter and dust wipes was 4.2 months (median 3 months); nearly half (14) of the units had been issued a lead hazard reduction letter in the past two months.

Housing units were grouped by building so that the EBL and non-EBL units’ dust levels could be compared.  If there were multiple non-EBL units in a building (for example, a four-unit building in which one unit had housed an EBL child), the dust lead levels for the multiple non-EBL units were averaged.  The average dust lead levels for the non-EBL units were compared to dust levels in the EBL unit.  There were 20 multifamily dwellings for which such comparisons were possible (11 of the units were single-family homes that had housed an EBL child).

           

Analysis

 

The average dust lead level for approximately a quarter of the units met the EPA standard for floors and window sills, with very few units passing the standard for window wells (Table 1).[1]  Only one of the units (a non-EBL unit) met the EPA standards on all three surfaces (floors, windows sills and window wells). Table 1 shows that EBL units were slightly less likely to pass the EPA floor dust standard than non-EBL units (23% EBL units versus 29% non-EBL units).  A t-test confirmed that there were no significant differences between the average dust levels in EBL and non-EBL units for floors, sills, or wells. 

To explore the effect of environmental investigation-ordered repair work on dust lead levels, the EBL units were compared to non-EBL units in the same building.    In approximately two-thirds of the buildings (Table 2), units that had received an environmental investigation had lower dust lead levels than non-EBL units.  In the remainder of multifamily units, the EBL units had the same or higher levels than the paired non-EBL units.   Thus, in the majority of units, the lead reduction work may have made units less hazardous in comparison with their paired non-EBL units, but most still remained well above EPA standards for lead in dust. 

 


 

Table 1: Comparison of average dust levels for floors, sills, and wells

 

FLOORS

Total number

Average dust lead (µg/ft2)

Median dust lead (µg/ft2)

Number under EPA standard

(40 µg/ft2)

Percentage units that meet the EPA standard

All units

66

192

73

17

26%

EBL units

31

194

87

7

23%

Non-EBL units

35

184

82

10

29%

 

SILLS

Total number

Average dust lead (µg/ft2)

Median dust lead (µg/ft2)

Number under EPA standard (250 µg/ft2)

Percentage units that meet the EPA standard

All units 

66

1188

536

15

23%

EBL units

31

1209

622

6

19%

Non-EBL units

35

536

554

9

29%

 

WELLS

Total number

Average dust lead (µg/ft2)

Median dust lead (µg/ft2)

Number under EPA standard (400 µg/ft2)

Percentage units that meet the EPA standard

All units

66

23384

9000

2

3%

EBL units

31

22327

11250

0

0%

Non-EBL units

35

24320

10412

2

6%

 

 

 

Table 2: Comparison of EBL units and non-EBL units in multifamily buildings

 

Multiple Unit Comparison

Floors

Sills

Wells

Number of EBL units with dust levels lower than non-EBL units

12

13

12

Percentage of EBL units with dust levels lower than non-EBL units

60%

65%

60%

 

Multiple Unit Comparison

Floors

Sills

Wells

Number of EBL units with dust levels same or higher than non-EBL units

8

7

8

Percentage of EBL units with dust levels same or higher than non-EBL units

40 %

 

35%

40%

 

In order to explore whether the time elapsed between issuance of the lead hazard reduction letter and the subsequent intake wipe test was correlated with measured lead levels, dust lead levels were plotted against time elapsed since the lead hazard reduction letter.  If the length of time since the investigation was completed was the primary determinant of dust lead levels, higher dust lead levels would be expected in the units in which lead hazard reduction letter was issued many months before the dust wipes were taken.  No such relationship was evident for floors, window sills, or window wells (Figure 1).

 


Figure 1. Dust lead levels versus time elapsed since lead hazard reduction letter issued
.

 

 

 

 

 


 


 

 

 

 

 

 


Discussion

 

            This data set provides us with a unique opportunity to compare paired units in the same buildings -- one that had and one that had not received environmental investigations by the county health department’s Lead Program.  Analysis of this data suggests that homes may not consistently meet EPA lead dust standards for clearance testing after environmental investigations by the Monroe County Health Department using the state framework.

            There are several factors that must be considered when interpreting this data.  First, the sample size is small.  However, since the program treated 31 EBL homes out of 191 EBL cases in Monroe County in 1998, the analysis represents 16% of all EBL cases from one year.  There is no reason to believe the units that entered the HUD grant were systematically different from typical EBL homes.  Thus, the EBL units in this data set is likely to generally represent the universe of units that poison children at levels above 20 µg/dL. 

Second, the assumption that units within the same building had similar paint and maintenance histories may not be true in all buildings.  On the other hand, given that based on initial dust wipes there were also significant lead hazards in the non-EBL units, this appears to be a reasonable assumption in general.

Third, some time had elapsed between the lead hazard reduction letter and dust lead measurements.  It is generally recommended that clearance testing be performed within hours after a final cleaning, not days or months.  On the other hand, HUD recommends reinspection of interim controls on an at least annual basis (1995 HUD guidelines, “reevaluation schedule”).  While interim controls are not expected to be permanent, these guidelines imply they should at least be effective for one year as long as there is regular monitoring, maintenance and cleaning.  Thus, it seems reasonable to expect that if the units were safe at the time of the lead hazard reduction letter, most of the units would still be safe after 3 months, the median time that had lapsed before samples were taken.

Fourth, not all properties screened for intake into the HUD program followed the same pattern of investigation and treatment.  For example, in one unit the EBL investigation did not address lead hazards in the attic, since that area was not used by the EBL child.  However, because the unit’s use patterns had changed by the start of the HUD program (the EBL child moved and the new family intended to use the attic), the HUD program risk assessment included dust wipes from the attic.  Thus, this unit had extremely high average dust levels, but did not necessarily pose a hazard to the child who had been living there.  Because the data was not generated for the purposes of this analysis and rigorously controlled for such anomalies, it is difficult to know what degree of uncertainty data problems introduce into this analysis. 

It is clear from Figure 1 that a small number of extremely high dust lead levels in the data set significantly increased the mean values of the dust wipes.  Given the potential data problems, one might conservatively assume that the outliers (very high dust levels) arose from such data anomalies.  This conservative assumption suggests that the medians may be more reliable than the means of the data set.  Comparison of the median values presented above is consistent with the findings of this analysis. 

This analysis strongly suggests that while units may be improved by county-ordered work as a result of state framework environmental investigations, we do not know whether or not residual lead levels in the dust are below those required by HUD to protect children’s health.  Without post-investigation dust wipes there is no measure of the resulting lead safety of the house.  It is even possible that dust lead levels have increased due to property owners’ failure to use lead safe work practices while making county-ordered repairs.  

In any case, the fact that no units in this analysis that had had an environmental investigation within the past 12 months met the EPA standards for lead dust on floors, sills and wells (although around a quarter of the units passed for one or more of those surfaces) strongly suggests that many of the units were not lead-safe upon completion of the work ordered under the state framework.   Further research on the dust lead levels in children’s homes following an EBL investigation could address the question of whether time elapsed between investigation and dust wipes samples in this program explain the high failure rate.  However, the graphs in Figure 1 suggest this is not the case.

The Monroe County Health Department carefully follows state guidelines in conducting EBL investigations.  However, since this protocol does not include sampling household dust for lead after county-ordered work is completed, it is not known whether homes that have been through an EBL investigation meet EPA lead dust standards.  The findings of this analysis suggest strongly that taking dust wipe tests after EBL investigation-ordered work is done is essential to make sure that these investigated units meet EPA lead dust standards at the time of completion.  This would improve the likelihood that EBL children do not continue to be exposed to lead hazards in their homes.

The Monroe County Health Department has a strong reputation in the state for effectively implementing the state lead program guidelines.  However, the state lead program does not require testing the lead in dust after work is completed as a result of EBL investigations.  For the reasons described above (including training workers, treating bare soil, clearance testing, and reevaluation), the HUD program is more protective of children’s health than is the state program.  This analysis suggests that the existing state guidelines do not consistently insure that the homes of Monroe County’s EBL children are made safe as measured by the EPA standard over time.  Implementing the HUD approach to lead hazard reduction is one way to increase the likelihood that EBL children do not live in homes that contain lead hazards.



[1] Note that the HUD standard actually applies to individual samples, whereas these are composites or averages of all floor dust samples in the unit.  However, this does give an overall sense of the dust lead levels in the unit.  In addition, we may certainly conclude that if the average fails the HUD standard, there is at least one individual floor where dust lead levels exceed the standard.