Navigating the Road to Recovery

Assessment of the Coordination, Communication, and Financing of the Disaster Case Management Pilot in Louisiana

by Joie D. Acosta, Anita Chandra, Kevin Carter Feeney

This Article

RAND Health Quarterly, 2011; 1(1):18

Abstract

In 2009, individuals heavily affected by Hurricanes Katrina and Rita were still in need of social services. The Federal Emergency Management Agency (FEMA) provided funding to the Louisiana Recovery Authority to implement the Disaster Case Management Pilot (DCMP) in order to help people still living in FEMA temporary housing units in April 2009 move to permanent and secure housing and access services. Despite concerted effort by participating agencies, the implementation of the DCMP was fraught with challenges. As a result, the pilot could not be implemented as intended, leaving the needs of many clients not fully met. This article shares details of a study that documents some of the key challenges in coordination, communication, and financing of the program and offers recommendations for future state and FEMA implementation of disaster case management. In light of these challenges, the authors recommend that federal and state governments review the systems used to identify and locate residents in need of disaster case management; these systems performed poorly in the DCMP, making it difficult to appropriately plan services. The stop and start of recovery initiatives led to serious discontinuities in client recovery, so the authors also recommend that federal and state governments consider a single, longer-term recovery initiative that seamlessly acknowledges the stages of human recovery. Improvements in how federal and state governments identify and locate affected residents, consider needs and vulnerabilities in planning, and ensure continuity of services are critical to ensure high-quality disaster case management.

For more information, see RAND TR-849-LRA at https://www.rand.org/pubs/technical_reports/TR849.html

Full Text

The impacts of Hurricanes Katrina, Rita, Ike, and Gustav continue to affect the Gulf States region. Thousands remain displaced from their homes and continue to struggle to recover from the trauma and aftermath. Historically marginalized and vulnerable populations in particular—such as individuals with disabilities, the elderly, and those from low socioeconomic backgrounds—confront barriers to recovery that others with more resources are able to resolve without the assistance of social services.

Disaster case management services provide relief to people in both the short and long term after disaster by connecting them with services needed to facilitate recovery. The Disaster Case Management Pilot (DCMP) is the most recent model of disaster case management for Louisiana and other states along the Gulf of Mexico, which was implemented by federal and state authorities in the period following Hurricanes Katrina, Rita, Ike, and Gustav. The Louisiana Recovery Authority (LRA) received funding from the Federal Emergency Management Agency (FEMA) to implement the DCMP (fall 2009–spring 2010) in order to fill gaps in service provision that still remained after earlier case management programs. The DCMP was specifically designed for individuals who still resided in FEMA temporary housing units as of April 27, 2009.

The LRA asked RAND to assess the DCMP. This analysis began in late March 2010 and included documentation of how the DCMP was organized and financed; identification of the major challenges to communication, coordination, and financing of the pilot; and provision of recommendations to the LRA and FEMA about how to improve future implementation of disaster case management in Louisiana. A team of RAND researchers used several methods, including document reviews, individual and group interviews with staff from the federal and state authorities responsible for implementing the pilot, focus groups with case managers and supervisors from the agencies contracted to provide case management, and analyses of case management data, to document DCMP activities and assess the pilot’s progress in helping residents of Louisiana obtain recovery services. The LRA and FEMA were interested in understanding optimal methods of disaster case management. Thus, the intent of this analysis is to identify implementation barriers and focus on areas for process improvement.

Implementation of the DCMP began in September 2009. During initial intake and triage of the 3,324 on the master list from FEMA, 722 clients were not able to be contacted due to out-of-date contact information, and 518 clients refused services. Between intake and clients being assessed by a permanently assigned case manager, DCMP case managers lost contact or were refused by another 280 clients. As of April 19, 2010*—one month before the pilot ended—Louisiana had approved invoices for less than half (44 percent) of the $9.4 million budgeted for the pilot and opened approximately half (n = 1,804) of the cases FEMA initially estimated. Among cases opened during the pilot period, 45 percent (n = 818) remained open as of April 2010, suggesting that these clients were still in need of case management services. Only 10 percent of the cases opened during the pilot (n = 186) were closed with at least one of the client’s primary needs met. An analysis of client characteristics found that most of these clients had multiple vulnerabilities: They were older (median age of 53); 82 percent had no more than a high-school education; more than 50 percent had an annual income of less than $15,000; and more than 75 percent of clients resided in a mandatory evacuation zone and were displaced from their primary residence. Statistical differences between clients with open and closed cases suggest that clients who had a recorded health issue were 41 percent more likely to still have open cases and that those who fell below the poverty line, had no source of income, or were otherwise unable to support themselves were 32 percent more likely to still have open cases. Most clients with open cases needed housing (62 percent), case management (56 percent), or furniture assistance (40 percent). Predisaster, 28 percent (n = 505) of the DCMP participants resided in New Orleans; the remaining were in surrounding areas.

Despite concerted effort by participating agencies, the implementation of the DCMP was fraught with challenges—most notably, difficulties that emerged from the particular vulnerabilities (e.g., age, disability, isolation) of the target population. A major barrier for the pilot was the overall design: The pilot was designed for individuals who were still struggling to move from FEMA trailers nearly five years after Hurricane Katrina, but, due to delays in the application process, the pilot period was only seven months long (September 2009–March 2010). Without significant planning and preparation, this was not a feasible timeline in which to serve this vulnerable population. In addition to timeline challenges for this vulnerable population, the lists of clients provided to states were not complete. This missing contact information presented difficulties in reaching eligible clients. These design challenges and the additional challenges summarized in Table 1 resulted in delays in services and financial reimbursement, tensions between the LRA and contractors regarding pilot implementation, and discrepancies between the number of cases initially estimated and the number of cases actually opened. As a result, the pilot could not be implemented as intended; now it has ended, leaving the needs of many clients not fully met.

Table 1

Challenges Experienced During Implementation of the Disaster Case Management Pilot in Louisiana

Aspect of Implementation Challenge

Structure of the pilot

The list of eligible cases that went to case management organizations from FEMA was out of date, resulting in overestimates of staff needs and poor allocation of resources based on client location.

The timeline for the pilot was not feasible in terms of start-up, planning, and transition, particularly given the vulnerabilities of the target population.

Narrow eligibility criteria missed some individuals still in need.

Lack of clarity on roles and responsibilities between the lead contractor and the LRA might have been exacerbated by the contracting structure and RFP process for the pilot.

Case managers and case manager supervisors reported minimal training on data entry and management and on the overall operational processes of the DCMP.

Case management organizations lacked operational capacity (e.g., management structure, phones, computers) to start immediately and lacked time and funds to build the organizational capacity needed.

There was a lack of community resources for client referrals.

Understaffing and inappropriate staffing at the LRA and lead contractor agency created challenges to implementing the pilot.

Communication among agencies involved in the pilot

The DCMP objectives for expected benchmarks and progress toward goal and vision were not clearly communicated.

Communication about the roles and responsibilities of each entity involved in the DCMP was inconsistent.

There was limited documentation of decisions associated with DCMP processes, and many of these decisions were poorly communicated to local contractors.

Communication problems resulted in delays in service decisions and financial reimbursement and changes in policies midstream.

Pilot financing

State officials and local case management agencies reported that difficulties emerged due to limited guidance on how to complete financial forms required by FEMA.

Clarity and timelines for reimbursement presented challenges, particularly for case management agencies. Reimbursement policies also did not align well with case management needs.

Guidelines for what could be included for indirect costs were confusing and resulted in financial loss for contracting agencies.

Given the short pilot duration and state regulations, there was no funding for pilot start-up.

Data collection and evaluation of the pilot

The LRA received regular updates of individual level client data, which created duplication in data entry and inefficiencies in tracking.

Data quality was questionable because case managers had difficulty entering data; quality assurance was also limited.

In light of these challenges, we recommend that, before implementing another disaster case management program in Louisiana, the state authorities, in partnership with local case management agencies should do the following:

  • Assess the needs of the population and available community resources to inform planning.
  • Revise the request-for-proposals (RFP) process used for disaster case management. The RFP for the lead contractor and the third-party evaluator should be released in advance of other RFPs so the contract can be awarded and materials can be prepared in advance of bringing case management agencies on board. The proposal should clearly state measurable goals and objectives and the roles and responsibilities of each agency. The RFP should also include start-up time to allow case management agencies to hire staff (including a qualified data-entry specialist), equip offices properly, and require ongoing training of case managers.
  • Develop state guidance for how to implement these types of grants with attention to financial procedures. A more streamlined process for invoice review that does not require multiple levels of review is also needed.
  • Create a common or centralized forum to share disaster case management templates (e.g., client forms, financial forms), guidance (e.g., directions for reimbursement), decisions related to design and implementation, and communication about resource availability (e.g., connecting case managers to identify available community services) for participating agencies to use throughout the program.

To improve development and implementation of a national disaster case management program, we recommend that FEMA take the following steps in designing a national disaster case management program:

  • Consider how to best track client information for vulnerable populations affected by disaster, and use predisaster data to identify “vulnerability hot spots.”
  • Develop a web-based knowledge center at program inception to provide centralized program information on an ongoing basis.
  • Create financial templates for state use that accommodate state variation in reimbursement and other contract requirements, and review responsibilities around reimbursement timelines. Financial templates should be revised to ensure that line items account for the needs and requirements of best practices in case management.
  • Consider how to best design a support system that can streamline intake and triage of cases and help determine client eligibility for services.
  • Target investments to maintain an ongoing infrastructure to support disaster case management, which might improve response time and save start-up costs.
  • Coordinate the transition points between individual assistance and disaster case management.

The DCMP also highlighted overarching questions about the processes and underlying principles of disaster case management. Research is needed to answer these questions. Addressing the following questions could help to improve how disaster case management is designed and implemented in the future:

  • What is the best way to identify and track the location of clients and client needs?
  • How can disaster case management programs be designed to best meet the needs of vulnerable populations in the immediate postdisaster period?
  • How can case management services best develop financial literacy among clients to ensure appropriate and responsible use of federal dollars?
  • How can state authorities identify, before a disaster, the local contractors and case management agencies that are best equipped to handle disaster case management?

Finally, this analysis highlights two themes critical for all recovery efforts. First, the system of identification and location of residents—particularly the populations most at risk due to preexisting and disaster-related events—is limited at best. Without a concerted review of these systems, government and case management agencies are unable to appropriately strategize for adequate service provision, including staffing algorithms, resource allocation, and development of a robust resource network. Second, the “stop and start” of recovery initiatives at both the federal and state levels might lead to serious discontinuities in client recovery. A single, longer-term recovery initiative that seamlessly acknowledges the stages of human recovery is merited.

Note: * Final data from May 14, 2010, provide a more updated summary of the number of open and closed cases but were not available in time for use in this report.

RAND Health Quarterly is produced by the RAND Corporation. ISSN 2162-8254.