Uses of HMIS to Support Disaster Operations and Recovery: Lessons Learned from Katrina/Rita

1
Uses of HMIS to Support Disaster Operations and
Recovery Lessons Learned from Katrina/Rita

• Brian Sokol, National HMIS TA Initiative, Abt
  Associates
• David Talbot, DSI Inc.
• Fran Ledger, Canavan Associates
• David Canavan, Canavan Associates (Facilitator)

2
Overview

• Learning Objectives
• Dynamics of Disaster Response
• Role of Information Management in Disasters
• HMIS Usage in 2-1-1 Disaster/Recovery Efforts in
  Louisiana
• City Of San Antonio uses of HMIS in Disaster
  Response

3
Learning Objectives

• Increased understanding of the possible functions
  of HMIS during and after a major disaster.
• Provide a broad cross section of obstacles
  encountered and successful interim and long-term
  solutions developed.

4
The Dynamics of Disaster Response

• Brian Sokol, Abt Associates

5
The Dynamics of Disaster Response

• Katrina and Rita
• The Context Homeless Services within the Total
  Disaster Response
• Stages of Disaster
• Proximity to Disaster
• Emergency Support Functions
• Respondents/Organizations in a Disaster
• Types of Respondents
• Components of Organization
• Mission Conflict
• Information Management and Data Coordination
• Disasters Disrupt the Response
• Planner vs. Academic Perspectives

6
Hurricanes Katrina and Rita
7
Eight Stages of Disaster Response

• Preparedness
• Planning
• Warning
• Response Day Of Impact
• Pre-Impact
• Post-Impact
• Recovery
• Restoration (six months)
• Reconstruction
• Mitigation
• Hazard Perceptions
• Adjustments

Cycles
Homeless Services
Where we are now
Based on Drabek, 1986
8
Spatial Dimensions
Organized Aid Life goes on as normal, except for
volunteerism
Filter Staging Ground for Relief Activities
Fringe Impact Some Damage
Homeless Services
Total Impact
Dynes, 1970
9
Mapping the Storms in Space and Time
Rita
Katrina
Source Louisiana Hurricane Impact Atlas
10
Aspects of Disaster Response Emergency Support
Functions

• Transportation
• Communications
• Public works and engineering
• Firefighting
• Emergency management
• Mass care, housing and human services
• Resource support
• Public health and medical services

• Urban search and rescue
• Oil and hazardous materials
• Agriculture and natural resources
• Energy
• Public safety and security
• Long-term community recovery and mitigation
• External affairs

Homeless Services
Dept of Homeland Security, National Response
Plan, 2004
11
ESF-6 Mass care, housing and human services

• Lead Agencies Red Cross and FEMA
• Tasks
• Sheltering
• Feeding
• Emergency first aid
• Providing information about victims
• Unifying families
• Bulk distribution
• Short-term and long-term housing assistance
• Crisis counseling
• Providing for special needs victims
• Processing benefit claims
• Delivering ice, water, and emergency commodities
• Mail service to affected areas

12
Respondents to a Disaster
Structure/Staffing
Walk-in Volunteers (Skilled/ Unskilled)
Tasks/ Mission
Homeless Services
Dynes, 1970
13
Components of Organization
Social Order More Coordinated
Collective Behavior More Flexible
Which comes first?

• Domain
• Formal recognition of authority
• Task
• Formal division of labor
• Resources
• Mobilized people and technologies
• Activities
• Specific behavior of people or groups

Kreps et al. 1986, 1989, 1993
14
Mission Conflicts within Shelters

• Crisis Management
• Return community to normalcy
• Focus on basic needs and operations
• Loose privacy and eligibility rules
• FEMA/Local Respondents/Red Cross (?)
• Residents are Evacuees
• Sample goal Close evacuation shelters quickly
• Human Services
• Restore individual lives to normalcy
• Overall Case Management
• Tight privacy and eligibility controls
• Victims/Human Service Organizations
• Residents are Homeless
• Sample goal Keep shelters open as needed

15
Information Management Processes

• Registration and Headcounts
• Who is in the shelter?
• Public Safety
• Ordering Supplies/Meals
• Tracking Vacancies/Reserving Beds
• Reunite Families
• Linking to Benefits and Services
• Coordinate Case Management

16
Axes of Data Coordination

• Intra-Organizational
• Same db for shelter registration, missing
  persons, case management, benefits receipt,
  health?
• Inter-Organizational
• Geographic Breadth
• One db for all evacuees across the country?
• Disaster/Non-Disaster Depth
• Use the same db (HMIS) for evacuation and
  homeless shelters?
• Coordination attempts can conflict with each
  other.
• E.g., coordinating between disaster shelters and
  the homeless system within a region may mean
  using a different disaster system than the rest
  of the country.
• Impossible to coordinate across all axes at once.
  

17
Model of Information Management in Disasters
Those most self-sufficient exit faster, creating
a shift in mission, activities, and appropriate
systems
Population Type
Conflict
Time
18
Disasters Disrupt the Response

• Emergency Management Perspective
• The goal should be for the victim to encounter
  one person who gathers all the necessary data and
  inputs it into a database that is shared and
  transparent among all human service providers at
  the Federal, State and local level as required.
  This will likely increase efficiency, reduce
  frustration of evacuees and expedite the delivery
  of services for eligible recipients.
• (The Federal Response To Hurricane Katrina
  Lessons Learned. White House Report, 2006)

19
Can this ever work in a Disaster?
Hypothesis If all these elements were addressed
in advance, then a comprehensive data system
could work.
20
Response to Rita at the Hirsch Shelter in
Shreveport

• Rita can be seen as a test case for the
  hypothesis. By the time Rita hit everything was
  up and running. What happened?
• Computers moved out to make room for cots
• 500 new people showed up, overwhelming intake
  processes handed paper forms
• Software experts left town, having lost their
  hotel rooms to evacuees
• Volunteers reverted to their comfort zone
• Rita hit Shreveport
• Water seeped through roof and floors
  (electrocution risk)
• Case mgmt volunteers re-tasked to sandbagging
  and trench digging
• Intermittent blackouts

21
Response to Rita at the Hirsch Shelter in
Shreveport
9/22 4PM
Same Room, 9/22 8PM
22
Disasters Disrupt the Response

• Unlike other emergencies, no clear distinction
  between incident and response, victim and helper
• Organizations have to respond to being directly
  impacted themselves (e.g., there can be direct
  and indirect loss of personnel, resources,
  equipment, and facilities) (Quarantelli, 1989)
• Disasters almost always create the wrong
  conditions and cut off planned resources
• Example NYCs Office of Emergency Management was
  destroyed on 9/11

23
Academic Perspective

• Much traditional disaster planning takes the
  disorganizing aspects of emergencyand attempts
  to achieve greater rationality and control of the
  anticipated situation...
• but the disorganizing aspects are necessary
  in order to develop the mobilization required to
  cope with the tasks at hand
• the end result is more rational and, in time,
  more efficient since a community has restructured
  itself to meet a set of problems which its
  previous structure could not.
• Disaster planning should be made in the context
  of these natural processes set off by a disaster
  event
• It should facilitate these processes, not
  impose a model of human and technological
  efficiency which has little relationship to
  reality. (Dynes, 1970)

24
Summary Thoughts

• Information management planners for disaster
  shelters should account for the following
• Shelters have evolving missions and diverse
  populations
• Evacuation Shelter Manage the crisis return
  to normalcy
• Homeless Shelter Provide long-term human
  services
• Order and coordination are often at odds with
  flexibility and adaptation disordered sometimes
  better
• Disasters disrupt the environment of disaster
  response.
• Avoid systems requiring expertise and
  infrastructure that are likely to fail under
  disaster conditions.

25
HMIS Usage in 2-1-1 Disaster/Recovery Efforts

• Fran Ledger,
• Canavan Associates

26

• It is not the strongest of the species that
  survives, nor the most intelligent, but the one
  most responsive to change.
• Charles Darwin

27
Louisianas HMIS and 2-1-1 History

• Nine HMIS Regions and Six 2-1-1 Regions
• All HMIS 2-1-1 Regions use same vendor product
• HMIS 2-1-1 has overlapping management
  software system
• 1 Software System 1 Agency
• New Orleans
• Shreveport
• Lake Charles
• 1 Software System 2 Agencies
• Baton Rouge
• Lafayette
• Monroe
• 2 Software Systems 2 Agencies
• Hammond
• Houma
• Alexandria

28
Key Components

• HMIS effectiveness will decrease with unclear
  leadership and limited controls
• Keys to disaster HMIS usage
• Clear chain of command
• Rapid quality assurance

29
Louisiana HMIS Response Timeline

• Friday, August 26th
• MOUs Activated
• UWNELA VIA LINK
• Office of Emergency Preparedness UWNELA
• Sunday, August 28th
• Mandatory Evacuation of New Orleans
• 1300 calls in first 12 hours
• VIA LINK staff relocating to UWNELA

30
Louisiana HMIS Response Timeline

• Monday, August 29th
• Katrina makes landfall
• Levee failures flood 80 of New Orleans,
• 1/3 of Louisiana housing damaged or destroyed
• Wednesday, August 31st
• New Orleans telephone landlines fail
• United Way of America with CenturyTel expands
  UWNELA call center

31
Louisiana HMIS Response Timeline

• Wednesday, August 31st
• 4 Hours 4 lines ? 56 lines
• All cell phone 2-1-1 calls route to UWNELA
• Call types roof top rescues ? clothing donation
• Thursday, September 1st
• 2500 calls/day
• UWNELA staff on 20 hr shifts
• United Way of America requests 2-1-1 volunteers
  from around the country on its 2-1-1 listserve
• HUD, first federal agency to offer National
  Technical Consultants and resources

32
Louisiana HMIS Response Timeline

• Saturday, September 3rd
• Arrival of National Call Center and Resource
  Manager Specialist
• Multi-state disaster HMIS launched
• Monday, September 5th
• 8,000 calls/day
• Mass need swamps FEMA Red Cross
• UNWELA houses and feeds 70 volunteers

33
Louisiana HMIS Response Timeline

• Saturday, September 17th
• 48 desktops arrive from HUD
• Call Center transitions from paper/pencil ? HMIS
• Saturday, September 24th
• HUD consultants on site
• 12,000 federal, state, and local resources made
  available
• Rita hits 310-Info evacuates to UWNELA
• 4,000 - 6,000 calls/day
• CNN, Red Cross FEMA publicize 2-1-1

34
Louisiana HMIS Response Timeline

• Tuesday, November 8th- 10th
• VIA LINK transitions back to New Orleans
• National Volunteers Depart
• 50,000 call sheets to be entered
• Today
• Regional 2-1-1s have revised MOUs
• Regional 2-1-1s developing single phone system
• Regional HMISs developing statewide HMIS

35

• Special Thanks to
• All the call center volunteers, local and
  national, for answering the call in our countries
  greatest tragedy!
• Janet Durden, Executive Director, United Way of
  Northeast Louisiana
• Marguerite Redwine, CEO, VIA LINK
• Karen Puckett, President, CenturyTel
• Peter Bishop, United Way of America
• Mike Roanhouse, Department of Housing and Urban
  Development
• Robert Bowman, President, Bowman Internet
  Systems, LLC
• Melissa Flourney, President, Louisiana
  Association of Non-Profit Organizations, LANO
• My Landlord on Elm St. for chairs to sit on and a
  rental to put them in!

36
Mass Shelters-City Of San AntonioSuccesses and
Lessons Learned in managing 17,000 refugees in
the city of San Antonio

• David Talbot
• VP of DevelopmentData Systems International

37
Facilities

• Multiple facilities, not all connected
• Kelly USA
• Former US Air Force Base
• Maze like-difficult to navigate
• Multiple massive buildings
• Freeman Coliseum
• Huge single area
• LEVI
• Huge single area

38
Prior To Arrival

• Around a 12 hour notice that the city would
  receive a fairly sizable number of evacuees.
• No one could quantify or give a vague range of
  numbers to expect.
• Guesstimates ranged between 2,000 and 40,000.
• The city immediately recognized this wasnt a
  short term disaster but a long term homeless
  problem and elected to use their HMIS system from
  day one.
• City began tapping resources, favors,
  beg/borrow/steal facilities, food, connectivity,
  medicine, volunteers.

39
Prior To Arrival-Continued

• Facilities were set up
• Designated medical areas, bathing, etc
• Shelter bed areas were divided into a grid.
• Red Cross intake forms underwent minor
  modification to add
• Wrist Band ID
• Assigned Section
• HMIS System was customized by the city to contain
  only the fields on the customized red cross
  intake.

40
A (fuzzy) look at the shelter
Grid Location
41
Evolution of HMIS usage-Phase 1

• Phase 1-Missing Persons
• Data Entry done by 200 concurrent untrained
  volunteers, constant rotation of new blood made
  training impossible.
• Some at local colleges working in two computer
  labs
• Keller-Williams Reality provided a good source of
  computer literate volunteers.
• System was customized to give the user two
  functions with a single path through the
  system-Find Clients/Add Client
• Rapid integration and collapsing of disparate
  systems.
• Excel spreadsheets/Access Databases/Other Systems
• Systems have a tendency to multiply like rabbits.
• Imported data from all systems in the greater San
  Antonio area, got all Katrina sites using the
  same system, then uploaded batch to Katrina safe.
• Wristbands and Grid Location in HMIS made it
  relatively easy to locate clients.
• Some normal system security measures were
  disabled (not all).

42
Evolution of HMIS usage-Phase 2

• Phase 2-Transition out as many people as possible
• New workgroup was created with expanded
  functionality specifically around the data points
  enabling a quick transition out of the shelter.
• Section 8 Enrollments in LASection 8 in Texas
• The Phase 1 super simple workgroup continued to
  operate for missing persons.
• Wristbands and Grid Location in HMIS made it
  relatively easy to locate clients.
• Normal system security measures re-enabled.
• Rita hit about this time, yielding a batch of
  short term evacuees.

43
Evolution of HMIS usage-Phase 3

• Phase 3-Case Management
• Virtually every evacuee was special needs of
  some sort- . Very old, disabled, children without
  parents
• System was expanded to full regular case
  management mode for case managers who began
  working with everyone in the shelter.
• During Phase 3, FEMA started bussing/flying in
  evacuees from other locations to consolidate
  efforts.
• Phase 3 transitioned into the Shaw Group, a
  subsidiary of Halliburton, that was contracted by
  FEMA to transition the remaining people out of
  shelters.
• They were given a workgroup and a group of user
  ids, the city was removed from management of the
  shelter.

44
In Retrospect

• What worked well
• Wrist Bands
• Grid Locations
• Remote Data Entry Sites
• Private business assistance (HEB, Keller
  Williams, Rackspace)

• What was a challenge
• Red Cross Intake Form
• Dozens of centralized Katrina databases on the
  internet.
• No clear chain of command. City? Red Cross? FEMA?
• Information Release problems
• Infrastructure

45
Next time around

• Recommended changes to intake forms
• Drop pre-disaster address
• Add birth date in addition to age
• Better family structure and a clearer definition
  as to if the family is together at shelter or
  not.
• A question Were you enrolled in any government
  assistance programs?
• Opt Out I do not want my information posted
  publicly.
• Chain of command is clearer and defined from day
  1?
• Localities handling evacuees use local HMIS
  system uploading to a single publicly searchable
  database via a defined standard such as HUD XML.

46
Questions

• For more information on these presentations
  please contact our presenters at
• Brian Sokol at brian_sokol_at_abtassoc.com
• Fran Ledger at fran_at_davidcanavan.com
• David Talbot at david_at_data-systems.com
• For more information on HUDs Disaster Technical
  Assistance Project, contact
• David Canavan at david_at_davidcanavan.com
• Ann Oliva at ann.oliva_at_gmail.com.