Appendix 5 - 2007 Protocol

Hazardous Substances Emergency Events Surveillance

(HSEES)

Protocol

March 2007

Division of Health Studies

Agency for Toxic Substances and Disease Registry

U.S. Department of Health and Human Services

Atlanta, Georgia 30333

Table of Contents

Introduction

General

Background

Goal and Objectives

Goal

Objectives

Methods

Introduction

Notification of a Reportable Event

Surveillance Definition

Information Needed to Achieve Objectives

Data Collection

Training Manual

Data Sources

Data Management

Data Entry

Timeliness and Completeness of Events

Quality Control and Assurance

Reliability

Data Analysis

Prevention Outreach

Dissemination of Results

Reference

Appendix A – Training Manual

Appendix B – Hardcopy Data Collection Form

Appendix D – Instructions for Prevention Outreach Plan and Logic Models

Appendix E – Initial Assessment Form for Prevention Outreach Activities

Appendix F – Data Clearance/Data Request Form

I. Introduction

A. General

The purpose of the protocol is to provide an overview of the functioning of the Hazardous Substances Emergency Events Surveillance (HSEES) system and the responsibilities of participating states and the Agency for Toxic Substances and Disease Registry (ATSDR). More detailed technical information can be found in the HSEES Training Manual (Appendix A), the HSEES Quick Reference Guide, the HSEES User Guide, Release Notes, and on the HSEES secure website.

B. Background

In an attempt to describe the morbidity and mortality experienced by employees, first responders, and the general public that result from acute hazardous substances releases, a surveillance system has been developed that is currently implemented in 14 states: Colorado, Florida, Iowa, Louisiana, Michigan, Minnesota, New Jersey, New York, North Carolina, Oregon, Texas, Utah, Washington, and Wisconsin.

The need for a state-based surveillance system for acute releases of hazardous substances was first suggested in a 1988 study on the sensitivity of three existing databases: the National Response Center (NRC), the Department of Transportation's (DOT) Hazardous Materials Information System (HMIS), and the Acute Hazardous Events Data Base (1). The study reported the shortcomings of any single national reporting source when trying to assess both the number and effects of hazardous substances emergencies. Of the 587 events reported to these three national databases during the study period, only eight (1%) appeared in all three systems.

The purpose of these databases was not to assess adverse health outcomes that result from hazardous substances emergencies, but rather to serve as a mechanism of enforcement or notifying other agencies (e.g., environmental, enforcement, commercial clean-up, insurance). The raw numbers collected by these databases do not describe the many variables that are associated with the morbidity and mortality resulting from these emergencies, nor do they stratify by populations (i.e., employees, responders, general population) affected. In addition, these databases are limited because of the number of events missed by failure of reporting as well as by the selective reporting of substances.

Barriers that have prevented establishing a surveillance system to assess the morbidity and mortality associated with acute releases of hazardous substances include: 1) no legal mandate requiring these data to be reported, 2) no standard definition of an acute release of a hazardous substance, 3) no standard definition as to when an acute release of a hazardous substance begins and when one ends, 4) defining the geographic area of concern, and 5) identifying the exposed population.

Data that are collected by federal (e.g., USEPA, DOT), state (e.g., departments of natural resources) and local agencies focus on the release of hazardous substances into, and the effects of these releases, on the environment as end-points. These data are not adequate to fully investigate the many risk factors associated with the morbidity and mortality resulting from the release of those hazardous substances.

Therefore, a surveillance system which focused on the public health, and not the environmental impact, of acute hazardous substances releases was established in 1989 by ATSDR. Five state health departments participated in the 1990–1992 pilot phase of the HSEES system. By 1995, 14 state health departments were participating in the surveillance system, and by 2001, the number of participating states increased to 16. Currently, there are 14 participating state health departments.

Consistent with ATSDR's mission to “serve the public by using the best science, taking responsive public health actions, and providing trusted health information to prevent harmful exposures and disease related to toxic substances,” ATSDR continues to make funds available for state health departments to participate in surveillance activities associated with acute releases of hazardous substances.

II. Goal and Objectives

A. Goal

The primary purpose of this program is to build capacity in state health departments to develop and maintain a state-based surveillance system for monitoring acute releases of hazardous substances and conducting appropriate prevention activities. The goal of this surveillance system is to provide data that can be used to reduce the morbidity and mortality resulting from acute releases of hazardous substances. Reductions in morbidity and mortality should occur in first responders, employees, and the general public.

B. Objectives

The objectives of the surveillance system are as follows:

1. To describe the distribution of acute releases of hazardous substances within the participating states, as well as nationally.

2. To describe the type and cause of morbidity and mortality experienced by employees, responders, and the general public as a result of selected acute hazardous substances releases.

3. Analyze and describe risk factors associated with the morbidity and mortality; and

4. Develop and propose strategies to reduce subsequent morbidity and mortality when comparable events occur in the future.

.

III. Methods

A. Introduction

The surveillance system is predicated on the ability of state agencies to document the total number of acute hazardous substances releases that occur within their state. Once the mandated state agency is notified, a means must exist to notify the state health department of the event (see "Notification of a Reportable Event"). Once the acute release of a hazardous substance is reported to the participating state health department, the components of the surveillance system consist of: 1) data collection, 2) data management, 3) data analysis, and 4) regular dissemination of reports to promote strategies to reduce morbidity and mortality.

B. Notification of a Reportable Event

State health departments should identify and develop memoranda of understanding with the state and local notification agencies in their states. In some states, this might entail one agency (e.g., state police or public safety). In other states, multiple agencies might be called in the event of a release or threatened release of hazardous substances. Since the primary focus of this project is to assess and potentially prevent acute adverse health effects, the state health departments need to be contacted within 48 hours of the release in order to have timely data entry (see Data Management section, 2. Timeliness and completion of events).

C. Surveillance Definition

Participating state health departments collect data on acute releases of hazardous substances that meet the case definition. The case definition is periodically revised to be responsive to emerging public health needs. Effective January 1, 2005, a HSEES event is defined as an uncontrolled or illegal acute release of any hazardous substance in any amount for substances listed on the HSEES Mandatory Chemical Reporting List maintained on the secure Website. For substances not on the list, events are included if the amount released is ≥ 10 lbs or 1 gallon. Threatened releases of qualifying amounts are included if the threat led to an action (e.g., evacuation) to protect the public health. The Petroleum Exclusion clause of the CERCLA legislation excludes any forms of petroleum that have not been refined to the point of becoming single-chemical products such as pure xylene (CERCLA/SARA). However, HSEES does record information about petroleum if it is released with a qualifying substance. A hazardous substance includes any substance which might reasonable be expected to cause adverse health outcomes..

D. Information Needed to Achieve Objectives

In order to accomplish the objectives, both environmental and health data need to be collected. Environmental data are needed to identify the substance or substances involved in the release. Health data are needed to identify the public health consequences (e.g., morbidity, mortality, and evacuations) associated with the event. The state health department must establish agreements with mandated or other appropriate notification agencies within their state to collect reports of hazardous substances releases. Information collected through these sources is used to complete all aspects of the web-based data entry system. Beginning in 2000, states were no longer required to fill-out a hardcopy of the data collection form (Appendix B), however, all supporting documentation needs to be retained for at least three years. The hardcopy data collection form is reviewed and approved by the Office of Management and Budget (OMB) every three years. Minor modifications to the form are made at the time of renewal. The current data collection form expires on 5/31/2008.

E. Data Collection

Data collected in the web-based data entry system include the following items:

• Event identification and notification information

• Time, date, and day of the week when the event occurred

• Event description including type of industry responsible for the event

• Event type (fixed-facility or transportation-related event)

• Geographical location and place within the facility where the event occurred

• Factors contributing to the release

• Substance, chemical, or trade name

• Specific information on injured persons: age, sex, type and extent of injuries, distance from spill, population group (employee, general public, responder, student), and type of protective equipment used

• Information about decontaminations, orders to evacuate or shelter-in-place

• Land use and population information to estimate the number of persons at home who were potentially exposed

• Response to and termination of the event

Complete investigation of an event may involve gathering information from several sources and entering the information from all sources into the HSEES web-based surveillance system. The web-based data entry system used by the states to collect the necessary data is designed to leave little room for misinterpretation. However, open fields have been provided for event-specific comments.

F. Training Manual

To further reduce the possibility of misinterpretation, a training manual for the web-based data entry system is provided to each participating state (Appendix A). The training manual is a quick reference to the types of responses which are acceptable and unacceptable in the HSEES system. The training manual provides explanations and instructions for each question in the data entry system. The training manual is provided to new states and to new personnel when joining the HSEES system. If the HSEES Coordinator is unable to answer a question after referring to the training manual, the coordinator should contact the ATSDR technical advisor.

G. Data Sources

Sources of information that the participating states can use to complete data entry include state/local police and fire departments, environmental agencies, various offices of emergency government, the manifest or inventory (depending on whether the incident is a transportation or a fixed-facility event), the incident commander's report, the National Transportation Safety Board (NTSB) report, the Red Cross (for evacuation center/shelter data), hospital admission and discharge records, information reported to the NRC, the media, emergency medical technicians' or ambulance reports, poison control centers, and interviews with witnesses and victims.

H. Data Management

Overall data management procedures are consistent with existing ATSDR policies and procedures. Participating state health departments are responsible for collecting HSEES data and entering it into the web-based application. ATSDR maintains one dataset consisting of the data from all participating states. Industry codes and chemicals released are chosen from a pick list and chemical categories are automatically assigned based on the chemicals entered. Data collected through the HSEES system consist of information on acute releases and not individuals. Therefore, no personal identifying information is collected.

1. Data Entry

For consistency in data collection and ease of analysis, ATSDR created a web-based data entry system as of January 1, 2000. The web-based application provides an easy way to enter data for qualifying acute releases into a centralized computer database. The advantages of the web-based system include 1) only one version of the data entry screen which facilitates changes, such as updating the substance pick list from which chemicals are chosen; 2) only one version of the dataset, so revisions/corrections only need to be made in one location; and 3) real-time data entry.

A training manual is also provided to each of the participating states. As the state health departments are notified of events that meet the case definition, they enter the data into the web-based data entry system. This assures that the "case counts" are accurate at any given time.

Each event has a unique, sequential identification (ID) number which is generated by the software each time a new event is entered into the system. New information about an event can be entered into the computer by re-opening a particular event ID.

2. Timeliness and Completion of Events

In order to be able to effectively promote and protect public health, all data available within 48 hours of the release needs to be entered into the system. ATSDR understands that there may be a few fields where the data are still unknown, e.g., what injury type every victim had, which can be entered later. Events should be entered as quickly as possible; even if the information available is incomplete. ATSDR also anticipates that each state will have no more than 40% of events not meeting the 48 hour data entry deadline. If more than 40% of events are not entered within 48 hours of the release, a concerted effort to continuously improve performance in the timeliness of entering information into the system is expected. States should document the reason for an event being entered greater than 48 hours after the event occurred.

Additionally, data entry for fixed-facility and transportation events should be completed within two months and four months of the date of occurrence, respectively. Again, ATSDR understands that there may be extenuating circumstances that prevent some events from being completed within these timeframes, however most events are expected to be completed within these timeframes.

3. Quality Control and Assurance

The states will retain documentation about the events. The web-based data entry system has built-in logical checks and range checks, as well as checks for valid values. ATSDR will perform checks for inconsistent or missing data periodically. Data checks will be conducted on a periodic basis with a two month lag time for fixed-facility events and four month lag time for transportation events. ATSDR will provide the states with a list of data items that are flagged as either inconsistent or as a critical missing field. States should check these flagged data items and make corrections in the system within one month of receiving the request.

Frequent data checks will improve completeness of the data and improve accuracy in reporting by allowing states to get all the information about an event while it is still fresh. Each state will review 10% of events quarterly to make sure data entry is accurate.

I. Reliability

As needed, ATSDR emails participating state health departments a “HSEES News Flash” that contains answers to questions submitted by coordinators so that all participating states will have access to the same information to maintain consistency in data entry. The News Flash also contains important information and decisions concerning HSEES.

J. Data Analysis

Analyses conducted by ATSDR include those using the combined data from all participating states as well as data specific to each participating state. These analyses include frequency distributions and cross tabulations of specific variables.

ATSDR provides annual reports describing the results of data collection from all states on the HSEES website. ATSDR generates annual reports approximately 6 months after the data are finalized. Annual reports consist of a series of tables and figures describing the results of data collection, in addition to a written summary of the data. In addition, ATSDR generates cumulative reports for specific blocks of time (i.e., 5 year data analysis).

The state health departments, by downloading their own state-specific datasets, may also prepare an annual report describing the results of data collection in their state for the specific year. The states may also perform descriptive and hypothesis testing analyses.

K. Prevention Outreach

In all participating states, the Project Coordinator should work with the Principal Investigator to develop strategies aimed at reducing morbidity and mortality associated with acute releases of hazardous substances. State health departments participating in HSEES for three years or more need to conduct at least three activities each calendar year; at least one of these activities needs to be counter-terrorism focused. Prevention outreach activities should be performed throughout the year, approximately quarterly. At least two activities need to be supported by state data (data-driven) and involve a mix of target groups and primary and secondary prevention, while the other activity can foster general awareness about the surveillance system. Instructions for formatting prevention outreach activities and corresponding logic models are presented in Appendix D. Prevention outreach plans and logic models for the following calendar year are due by December 31 of the current calendar year.

For state health departments who have participated in HSEES for less than three years, the schedule for conducting prevention outreach activities is as follows:

• 1^st year: conduct one activity to foster general awareness about the surveillance system

• 2^nd year: conduct one awareness activity and one data-driven activity

• 3^rd year and beyond: conduct one awareness activity and two data-driven activities

Additionally, all participating state health departments need to document instances when they alerted appropriate public health partners about hazardous substance events with potential human health hazards that need immediate follow-up action. States are also to record any follow-up activities that resulted from the alert. Participating state health departments will alert appropriate public health partners on at least 5 events/year within 48 hours of event occurrence and submit a quarterly alert log to ATSDR which also needs to include any instances of policy, legislation, or planning influenced by HSEES.

Tracking the audience numbers and evaluations of the activities is useful in demonstrating the system’s effectiveness. Within two weeks of completing a prevention outreach activity, Project Coordinators will need to submit an Initial Assessment Form (Appendix E ) stating the actual audience number (i.e., the number of fact sheets distributed, the number of attendees at the conference session) and any anecdotal information or feedback about the activity. If ATSDR does not receive the Initial Assessment Form within two weeks, the participating state health department will be notified. ATDSR will also implement a new tracking system to more formally track evaluations of data-driven activities. The evaluations of data-driven activities will be linked to the effectiveness measures proposed in the prevention outreach plans (i.e., comparing the number of events and/or victims for a specified time period prior to the presentation/distribution of material to a specified time period after the presentation/distribution to see if the numbers decreased).

A summary of the initial assessments of the current calendar year’s prevention outreach activities is due by December 31. For prevention outreach activities that can not be completed in the current calendar year, participating state health departments need to decide by December 15^th if they wish to 1) request a one-month extension to complete the activity, 2) move the activity to the following year’s prevention outreach plan, or 3) drop the activity.

If requested, ATSDR will assist the states in accessing other states’ data or all states data to strengthen analyses for prevention outreach and identify similar types of events that are occurring regionally or system-wide; however, participating states do not routinely have access to the raw data from any other state.

Examples of awareness activities include distributing annual reports and presenting a general overview of the system to LEPCs, first responders and other interested parties. Examples of data-driven activities include analyzing county- or zip code-specific data for presentation to LEPCs; developing and distributing fact sheets on the most frequently released chemicals; providing HSEES data and collaborating with other state agencies to reduce mercury spills or meth lab events; and conducting a more intense evaluation of a previous year’s activity.

Data clearance forms for prevention outreach activities need to be submitted by the Principal Investigators (Appendix F) along with the material to be reviewed for approval by the ATSDR technical project advisor. ATSDR will attempt to review these materials as quickly as possible, but each review of a draft may take up to two weeks, and more than one draft may be necessary. Therefore, states should allow a month for the overall review process.

L. Dissemination of Results

Annual reports from the full dataset (all participating states) are posted on the ATSDR HSEES website approximately 6 months after the data are finalized. Annual reports from 1995 on are available on the ATSDR HSEES website.

States are encouraged to disseminate their data to those who can use it for prevention activities as well as to use their data for publications/presentations to appropriate audiences. Any data that is presented outside of the health department requires prior ATSDR approval before release or presentation. This includes quarterly, annual, and cumulative reports that will be released to the public or any agency, organization, or group outside of the health department (besides ATSDR); fact sheets, graphs, tables, maps, and other compilations of data; posters, slides, and any written text of oral presentations; articles for publication in state or private newsletters, magazines, or journals; and analyses conducted by third parties, such as universities. If it is necessary to release raw data to an entity outside of the HSEES program, a data sharing agreement should be completed. Instructions on how to do this can be obtained from the ATSDR technical advisor.

Principal Investigators need to complete a data clearance form (Appendix F) and submit it with the material to be reviewed for approval by the ATSDR technical project advisor. ATSDR will attempt to review these materials as quickly as possible, but each review of a draft may take up to two weeks, and more than one draft may be necessary. Therefore, states should allow a month for the overall review process.

IV. Reference

1. Binder S. Deaths, injuries and evacuations from acute hazardous materials releases. Am J Public Health 1989; 79: 1042-1044.