10

PROPOSED DATA ANALYSES

image

As the committee members read the legislation that directed their charge and reviewed the literature, they recognized that not all areas of concern were addressed by available information. Thus, they conceived a detailed data-analysis plan that they believed would provide clarity on many of the issues of concern. Initially, the committee thought it might carry out the proposed analyses and researched sources of potential data, developed data-analysis plans, data use agreements, and provided the institutional review boards of the National Academies, Department of Defense (DOD), and Department of Veterans Affairs (VA) with the appropriate documents. The committee abandoned its plans because of time and funding constraints, however, this chapter describes the committee’s efforts for exploring and gathering data for the development of a database that would provide the foundation for linking data from multiple data sources (e.g., linking basic demographic data with data on diagnoses, types of treatments received, and employment and earnings records before and after deployment). The chapter details the committee’s vision for the database, the questions the committee hoped to address with the data, the data sources the committee believed would provide answers to its questions, and the committee’s overall process for approaching this task.

Through that process, the committee identified a variety of data and databases available from the DOD, VA, and other federal agencies that could be used to address many of the questions posed by the legislation that motivated its work. Appendix F describes a small sample of the databases explored by the committee as the list of potentially useful data sources continued to increase with additional queries and searches.

On the basis of its investigations, the committee determined that it should expand its initial approach in responding to its task by proposing data analyses that could be conducted by using existing government databases. Very limited quantitative characterization of the issues described in the legislation existed in the published literature, so the committee proposed using available government data and linking them. There are no databases or files that fully integrate basic deployment and demographic data with health outcomes, treatment or transition of care, access to care, employment and earnings records before and after deployment, family records, and other processes and outcomes. Thus, the committee decided to develop an analytic database for that purpose and attempted to begin linking the data. As noted, the committee ultimately abandoned its approach, however, it details its efforts so that others might carry out its vision.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 457
10 PROPOSED DATA ANALYSES As the committee members read the legislation that directed their charge and reviewed the literature, they recognized that not all areas of concern were addressed by available information. Thus, they conceived a detailed data-analysis plan that they believed would provide clarity on many of the issues of concern. Initially, the committee thought it might carry out the proposed analyses and researched sources of potential data, developed data-analysis plans, data use agreements, and provided the institutional review boards of the National Academies, Department of Defense (DOD), and Department of Veterans Affairs (VA) with the appropriate documents. The committee abandoned its plans because of time and funding constraints, however, this chapter describes the committee’s efforts for exploring and gathering data for the development of a database that would provide the foundation for linking data from multiple data sources (e.g., linking basic demographic data with data on diagnoses, types of treatments received, and employment and earnings records before and after deployment). The chapter details the committee’s vision for the database, the questions the committee hoped to address with the data, the data sources the committee believed would provide answers to its questions, and the committee’s overall process for approaching this task. Through that process, the committee identified a variety of data and databases available from the DOD, VA, and other federal agencies that could be used to address many of the questions posed by the legislation that motivated its work. Appendix F describes a small sample of the databases explored by the committee as the list of potentially useful data sources continued to increase with additional queries and searches. On the basis of its investigations, the committee determined that it should expand its initial approach in responding to its task by proposing data analyses that could be conducted by using existing government databases. Very limited quantitative characterization of the issues described in the legislation existed in the published literature, so the committee proposed using available government data and linking them. There are no databases or files that fully integrate basic deployment and demographic data with health outcomes, treatment or transition of care, access to care, employment and earnings records before and after deployment, family records, and other processes and outcomes. Thus, the committee decided to develop an analytic database for that purpose and attempted to begin linking the data. As noted, the committee ultimately abandoned its approach, however, it details its efforts so that others might carry out its vision. 457

OCR for page 457
458 RETURNING HOME FROM IRAQ AND AFGHANISTAN COMMITTEE’S APPROACH First, the committee reviewed the legislation to determine which sections could be addressed with data and which could not; it then divided each section into discrete tasks that would be overseen by the committee’s work groups. Second, each work group developed an analytic plan that highlighted the questions to be answered and the data sources that could potentially provide the necessary information; thus, the issues and questions posed were aligned with available data sources. Third, the committee prepared its analytic plans for the institutional review board (IRB) at the National Academy of Sciences and for data-use agreements (DUAs) with various government agencies. Fourth, the work groups attempted to secure the necessary data; but they were often faced with impediments that prevented accessing them. Finally, the committee recognized that not all the questions that it would like to have answered would have a data source and that some questions would remain unanswered. COMMITTEE PROCESS The need and request for each specific file and variable within a file from various data sets would be supported by different sections of the legislation (see Appendix A). Because the legislation was broad and covered so many aspects of deployment—not only in active duty but in veterans, family members, and communities—the committee members reviewed the legislation carefully and developed questions that would need to be answered in an effort to respond to each section of it. The committee members also had to determine which agencies collected the information that might enable them to answer the questions and which datasets would provide the data that they needed. Data dictionaries were secured and reviewed so that the committee members were able to identify the files that they needed and to determine whether the information was held in a particular database. On the basis of the legislation, several data groups were organized according to the need for similar data (rather than according to the committee’s work groups). For example, the outcomes, treatment, family, and economic work groups wished to understand how the number of deployments affected outcomes of the conditions of interest, such as posttraumatic stress disorder (PTSD) and traumatic brain injury (TBI). The committee members planned to send the Defense Manpower Data Center (DMDC) file of all deployed persons to the VA and DOD health systems to identify those who had the International Classification of Diseases (ICD) diagnoses of interest. Those persons would then be followed through the DOD and VA health system to determine the types of treatment provided and the numbers of visits to treatment facilities, they would then be linked to employment records and earnings so that the committee could determine the impact of deployment. The committee secured the demographic files of all those deployed (see Chapters 2 and 3), but linking them with the additional files needed became too time-consuming and the committee would not be able to complete its work before the termination of its contract with the DOD. DATA-COLLECTION PLAN AND ANALYSIS QUESTIONS The following is a breakdown of the legislation according to the types of data needed, samples of the types of questions developed by each group (to be answered by the data), and the data sources that the committee believed would provide answers to the questions. In some cases,

OCR for page 457
PROPOSED DATA ANALYSES 459 more than one work group examined the same section of the legislation, but each group had its own focus. Public Law 110-181. Section 1661. Study on physical and mental health and other readjustment needs of members and former members of the armed forces who deployed in Operation Iraqi Freedom and Operation Enduring Freedom and their Families. (2) A second phase, to be completed not later than three years after the date of the enactment of this Act, to carry out a comprehensive assessment, in accordance with the parameters identified under the preliminary report required by paragraph (1), of the physical and mental health and other readjustment needs of members and former members of the Armed Forces who deployed in Operation Iraqi Freedom or Operation Enduring Freedom and their families as a result of such deployment, including, at a minimum— (c) POPULATIONS TO BE STUDIED.—the study required under subsection (a) shall consider the readjustment needs of each population of individuals as follows: (1) Members of the regular components of the Armed Forces who are returning, or have returned, to the United States from deployment in Operation Iraqi Freedom or Operation Enduring Freedom. (2) Members of the National Guard and Reserve who are returning, or have returned, to the United States from deployment in Operation Iraqi Freedom or Operation Enduring Freedom. (3) Veterans of Operation Iraqi Freedom or Operation Enduring Freedom. (4) Family members of the members and veterans described in paragraphs (1) through (3). Data Group 1 Relevant Sections of the Law (B) an assessment of the particular impacts of multiple deployments in Operation Iraqi Freedom or Operation Enduring Freedom on such members and former members and their families; (C) an assessment of the full scope of the neurological, psychiatric, and psychological effects of traumatic brain injury on members and former members of the Armed Forces, including the effects of such effects on the family members of such members and former members, and an assessment of the efficacy of current treatment approaches for traumatic brain injury in the United States and the efficacy of screenings and treatment approaches for traumatic brain injury within the Department of Defense and the Department of Veterans Affairs; (D) an assessment of the effects of undiagnosed injuries such as post-traumatic stress disorder and traumatic brain injury.

OCR for page 457
460 RETURNING HOME FROM IRAQ AND AFGHANISTAN Data Files Required DMDC ID data VA DMDC Dx with TriCare DMDC DMDC PDHA & dates Dx with Demographic Deployment PDHRA PTF & OPC & dates data C&P Sample Questions to Be Answered by the Data  How many people who were deployed screen positive for TBI, PTSD, or both? How many people who were deployed receive a diagnosis of TBI, PTSD, or both (in theater vs posttheater)? How long after deployment do service members and veterans receive a diagnosis of PTSD, TBI, or both?  What are the major comorbidities associated with TBI, PTSD, depression, substance use, and suicidal ideation?  What is the impact of multiple deployments, length of deployment, and site of deployment on TBI, PTSD, or both?  How do the above questions vary by sex, ethnicity, marital status, age, service, rank, education attained, criminal record, home of record on service entry, aptitude, occupational category, branch of service, reserve, guard, or active duty?  Does higher education affect TBI or PTSD outcomes? Does alcohol use before deployment increase or decrease the risk of TBI or PTSD?  In a longitudinal cohort of deployed service members, what are the frequency, course (timing), and co-occurrence of screening for, diagnosis of, and treatment for TBI, PTSD, substance-use disorder, depression, and suicide risk?  When treatment is provided for one of the five conditions (as indicated by ICD-9 code), what approach is used? That is, if pharmacologic, what medication? If behavioral, what method, by whom, and were sessions individual or group?  What is the person’s current functioning: What diagnoses is the person still being treated for? What is the person’s current employment status? Quality of life?  Does the course of screening, diagnosis, and treatment vary by branch of military, where processed through the immediate postdeployment period, ZIP code of postdeployment residence, urban vs rural setting, rank, race or ethnicity, sex, marital status, waivers granted to allow enlistment, Armed Forces Qualification Test (AFQT) or General Classiification Test, pre-enlistment disorders, or predeployment disorders?  What are the number of deployments and the length of deployments of each person? What are the number and pattern of medical visits? What proportion of people who have a particular diagnosis are using services in DOD or in VA? Data Group 2 Relevant Sections of the Law (A) an assessment of the psychological, social, and economic impacts of such deployment on such members and former members and their families;

OCR for page 457
PROPOSED DATA ANALYSES 461 (B) an assessment of the particular impacts of multiple deployments in Operation Iraqi Freedom or Operation Enduring Freedom on such members and former members and their families; (C) an assessment of the full scope of the neurological, psychiatric, and psychological effects of traumatic brain injury on members and former members of the Armed Forces, including the effects of such effects on the family members of such members and former members, and an assessment of the efficacy of current treatment approaches for traumatic brain injury in the United States and the efficacy of screenings and treatment approaches for traumatic brain injury within the Department of Defense and the Department of Veterans Affairs; (I) an assessment of the impacts of increasing numbers of older and married members of the Armed Forces on readjustment requirements. Data Files Required DMDC DMDC VA TriCare Entrance Demographic Dx Dx medical data exam DMDC ID data DMDC VA NDI DMDC TriCare Vital status Vital status Cause of Deployment Vital status data data death Sample Questions to Be Answered by the Data  Is mortality higher on return from deployment than among the nondeployed?  What are the effects of multiple deployments, length of deployment, and location of deployment?  Does TBI or PTSD increase mortality? Is a person more likely to be in a motor-vehicle accident within 2 years after a TBI or PTSD diagnosis than a person who does not have a TBI or PTSD diagnosis?  How do answers to the above questions vary by sex, ethnicity, marital status, age, service, rank, educational attainment, criminal record, home of record on service entry, aptitude, occupational category, or branch of service?  Do people who have TBI or PTSD who are deployed have a higher risk of being killed or injured in action than others who are deployed? How many people who have TBI or PTSD complete their deployment? Data Group 3 Relevant Sections of the Law (A) an assessment of the psychological, social, and economic impacts of such deployment on such members and former members and their families;

OCR for page 457
462 RETURNING HOME FROM IRAQ AND AFGHANISTAN (B) an assessment of the particular impacts of multiple deployments in Operation Iraqi Freedom or Operation Enduring Freedom on such members and former members and their families; (C) an assessment of the full scope of the neurological, psychiatric, and psychological effects of traumatic brain injury on members and former members of the Armed Forces, including the effects of such effects on the family members of such members and former members, and an assessment of the efficacy of current treatment approaches for traumatic brain injury in the United States and the efficacy of screenings and treatment approaches for traumatic brain injury within the Department of Defense and the Department of Veterans Affairs; (E) an assessment of the gender- and ethnic group-specific needs and concerns of members of the Armed Forces and veterans. Data Files Required DMDC ID data DMDC DMDC VA VA TriCare Demographic Deployment Rx Vital status Rx data Sample Questions to Be Answered by the Data  How many people are deployed or redeployed while on antidepressants, antianxiety agents, or narcotic pain medications?  How does the answer to the above question vary by sex, ethnicity, marital status, age, service, rank, education attained, criminal record, home of record on service entry, aptitude, occupational category, or branch of service? For example, the outcomes, treatment, family, and economic groups wished to understand how the number of deployments affected positive screens and subsequent diagnoses with the conditions of interest (e.g., PTSD or TBI)?  Are people who report antidepressant or antianxiety medication use before deployment at increased or decreased risk for TBI and PTSD?  Are deployed people who are using antianxiety or antidepressant medication at higher or lower risk for injury or death than those not taking the drugs?  When treatment was provided for one of the five conditions (as indicated by ICD-9 code), what approach was used? That is, if pharmacologic, what medication? If behavioral, what method, by whom, and were sessions individual or group? Data Group 4 Relevant Sections of the Law (A) an assessment of the psychological, social, and economic impacts of such deployment on such members and former members and their families;

OCR for page 457
PROPOSED DATA ANALYSES 463 (B) an assessment of the particular impacts of multiple deployments in Operation Iraqi Freedom or Operation Enduring Freedom on such members and former members and their families; (D) an assessment of the effects of undiagnosed injuries such as post-traumatic stress disorder and traumatic brain injury, an estimate of the long-term costs associated with such injuries, and an assessment of the efficacy of screenings and treatment approaches for post-traumatic stress disorder and other mental health conditions within the Department of Defense and Department of Veterans Affairs; (E) an assessment of the gender- and ethnic group-specific needs and concerns of members of the Armed Forces and veterans. Data Files Required NCIC UCMJ NDI Criminal Criminal COD record record DMDC ID data DMDC DOD VA DMDC TriCare Deployment Demographic Sexual Dx & date Dx & date data trauma Sample Questions to Be Answered by the Data  How many people report sexual trauma or harassment during deployment?  What are long-term outcomes of sexual trauma (in men and women) during deployment?  Does sexual trauma increase the risk of PTSD, TBI, or both?  Does TBI or PTSD increase the risk of sexual trauma among women and men?  Does TBI or PTSD diagnosis increase the risk of becoming a victim of domestic abuse or of committing domestic abuse?  Do education, age, sex, race, ethnicity, service branch and component, residence of origin, number and length of deployments, and marital status increase the risk of domestic abuse in people diagnosed with TBI, PTSD, or both? Data Group 5 Relevant Sections of the Law (A) an assessment of the psychological, social, and economic impacts of such deployment on such members and former members and their families;

OCR for page 457
464 RETURNING HOME FROM IRAQ AND AFGHANISTAN (B) an assessment of the particular impacts of multiple deployments in Operation Iraqi Freedom or Operation Enduring Freedom on such members and former members and their families; (D) an assessment of the effects of undiagnosed injuries such as post-traumatic stress disorder and traumatic brain injury, an estimate of the long-term costs associated with such injuries, and an assessment of the efficacy of screenings and treatment approaches for post-traumatic stress disorder and other mental health conditions within the Department of Defense and Department of Veterans Affairs; (H) an assessment of the impacts on communities with high populations of military families, including military housing communities and townships with deployed members of the National Guard and Reserve, of deployments associated with Operation Iraqi Freedom and Operation Enduring Freedom, and an assessment of the efficacy of programs that address community outreach and education concerning military deployments of community residents; (I) an assessment of the impacts of increasing numbers of older and married members of the Armed Forces on readjustment requirements. Data Files Required HRSA UCMJ Homelessness Criminal data record DMDC ID data CDC BRFSS DMDC DMDC NCIC DMDC SSA Demographic Marital Criminal Deployment Income Used for trend analysis data status record Sample Questions to Be Answered by the Data  How is quality of life affected by PTSD or TBI? Professionally? Personally? Domestically?  How many of those who have TBI or PTSD are unemployed? Incarcerated? Divorced?  Does TBI or PTSD diagnosis affect income, employment, or vocational needs (such as the need to change profession)?  Do such factors as education, age, sex, race, ethnicity, service branch and component, residence of origin, number and length of deployments and marital status increase the risk of homelessness in people who have TBI, PTSD, or both?  What is the person’s current functioning? What diagnoses are still being treated for? What is the current employment status? What is the current quality of life?  What are the longitudinal effects of deployment to OEF and OIF on active-duty personnel and recent veterans with respect to earnings? Military earning? Employment? Homelessness? Arrest?

OCR for page 457
PROPOSED DATA ANALYSES 465  How have changes in the benefits provided under the GI Bill affected educational attainment? How might they affect earnings (as a consequence of increased educational attainment)? How are the effects this moderated by deployment?  What are the economic effects of deployment on spouses in particular and families in general? Data Group 6 Relevant Section of the Law (C) an assessment of the full scope of the neurological, psychiatric, and psychological effects of traumatic brain injury on members and former members of the Armed Forces, including the effects of such effects on the family members of such members and former members, and an assessment of the efficacy of current treatment approaches for traumatic brain injury in the United States and the efficacy of screenings and treatment approaches for traumatic brain injury within the Department of Defense and the Department of Veterans Affairs. Data Files Required DMDC DMDC Demographic Deployment data DMDC ID data VBA VHA TriCare C&P OPC/PTF Procedures Procedures Sample Questions to Be Answered by the Data  How many people who receive diagnoses of TBI, PTSD, or both receive military or VA disability compensation? How many receive vocational rehabilitation services?  How frequently was imaging used to assist in diagnosis of TBI? Data Group 7 Relevant Sections of the Law (A) an assessment of the psychological, social, and economic impacts of such deployment on such members and former members and their families; (B) an assessment of the particular impacts of multiple deployments in Operation Iraqi Freedom or Operation Enduring Freedom on such members and former members and their families; (F) an assessment of the particular needs and concerns of children of members of the Armed Forces, taking into account differing age groups, impacts on development and education, and the mental and emotional well-being of children.

OCR for page 457
466 RETURNING HOME FROM IRAQ AND AFGHANISTAN Data Files Required DEERS SSA Family Income members DMDC ID data TriCare NCIC UCMJ DMDC DMDC Dx with Criminal Criminal Demographic Deployment dates record record data Sample Questions to Be Answered by the Data  What are the implications for children’s (and other family members’) well-being and functioning of each of the following characteristics: deployment frequency and duration, dwell time, location, type of mission, parental injury or death and the aftermath, and perinatal stressors (such as low birth weight and maternal smoking and drinking)?  What characteristics contribute to military families’ instability, such as divorce, parental estrangement, young parental age, parental unemployment or underemployment, and parental educational level?  What are the family effects of family income, child support, parental mental illness, parental drug abuse, parental crime, Post-Deployment Health Assessment (PDHA) or Post- Deployment Health Reassessment (PDHRA) findings, use of health services, diagnosed injury or illnesses, or previous trauma, illness, psychiatric diagnoses, and treatment?  How should military families be defined—such as by using deployment data, demographic data, rank, grade, income, military occupational specialty, AFQT score, or education level? Thus, the committee read the legislation, decided on questions that needed to be answered, and identified data sources. After reviewing the questions, which were organized by data groups, and examining potential data sources, the committee planned to link the data into a central database that was developed specifically for it. Figures 10.1 and 10.2 illustrate the committee’s vision for data-gathering and data inputs. The committee expected to conduct analyses to answer its questions and to shed light on the issues posed by the legislation. Other groups undertaking such analyses will likely develop their own view of database development and data inputs. The committee proposed to incorporate the details on all those (including active-duty, reserve, and National Guard personnel) deployed to OEF, OIF, and OND during 2001–2010 and to compare them with a matched cohort of personnel who were not deployed during that period. The database would include files provided by DOD, for example, by the DMDC (demographic data), TRICARE, the Defense Enrollment Eligibility Reporting System, the PDHA, and the PDHRA; by VA, specifically, by the Veterans Benefit Administration and Veterans Health Administration, such as Patient Treatment Files, Outpatient Clinic Visits, and Compensation and Pension; by the Centers for Disease Control and Prevention, such as the National Death Index and the Behavioral Risk Factor Surveillance System; by the Health Resources and Services Administration (data on homelessness); by the Department of Health and Human Services

OCR for page 457
PROPOSED DATA ANALYSES 467 (Medicare and Medicaid data); by the Social Security Administration; by the Internal Revenue Service; and by the Federal Bureau of Investigation (crime data) (see Appendix F for database descriptions). Initially, the committee requested the DMDC demographic database on all service members (regular, reserve, and National Guard Components) who were deployed to OEF and OIF during the period September 11, 2001–December 31, 2010, and those who were not deployed during the same period. In addition, the committee requested the available PDHA and PDHRA data files on deployed personnel. The DMDC files constituted the sampling frame from which multiple cohorts would be selected—initially with deidentified files that contained no names and no Social Security numbers and then, after selection, with unique person identifiers that permitted linkage to records from the different DMDC databases. Veterans Health Veterans Benefit Health Resources Uniform code of R135 AdministraƟon AdministraƟon (VBA VBA) and Services CMS Military JusƟce OEF & OIF Readjustment Data (VHA VHA) Acquire Fact of Death AdministraƟon Acquire Medicaid / (UCMJ ) UCMJ Requirements Acquire OPC / PTF and a Acquire Medicare endpoints Acquire criminal Registry Data , etc. variety of endpoints homelessness data records Legend Process NaƟonal Criminal Tricare and Defense Eligibility NaƟonal Death Social Security InformaƟon Center Subprocess Internal Revenue Enrollment ReporƟng System Index (NDI NDI) AdministraƟon (NCIC ) NCIC) Service (IRS IRS) (DEERS ) DEERS Acquire fact & cause (SSA ) SSA Federal Bureau of Decision Acquire address Acquire Dx & Rx, etc. of death (COD) Acquire income InvesƟgaƟon (FBI FBI) Acquire criminal records Database Yellow Databases can link to unique ID but not SSN Red Database can link to Yes Center for Disease both unique ID & SSN Control (CDC) CDC) Acquire BRFSS Tricare for Dx & Rx of R135 ID Master interest No Selected Study Request and Get Real ID required DEERS for Real ID to IOM Cohort for each Access to Data for matching ? Dependents & group’s use Spouses ID DMDC Real ID ID Sexual DMDC Real ID conversion Harassment conversion Start Here Defense Manpower Is subject Data Center (DMDC) R135 Determine Study included in DOD Data Request Demographics from Selected Study Demographic Cohorts & selected study Yes Access Granted 2001-2010 for Cohort ID to DMDC Master Controls requiring more OEF & OIF & follow -up? Non-deployed Match ID to Post Deployment Health Is subject data R135 Assessment (PDHA) & Post Provide & update In PDHA or Yes PDHA/PDHRA Deployment Health Reassessment Survey Data PDHRA file ? Master (PDHRA ) for Deployed (& Non-deployed ?) Provide Military medical and R135 apƟtude data from the Military Entrance Exam for Entrance Exam Deployed & Non-Deployed Master FIGURE 10.1 Committee’s database and data requirements.

OCR for page 457
468 RETURNING HOME FROM IRAQ AND AFGHANISTAN Process to idenƟfy subjects with R135 Study Dx or CondiƟons of Interest PopulaƟon of Deployed and Red Database can link to Non-deployed both unique ID & SSN Subjects with No Dx No Dx of interest? Yes of Interest Yes Subjects Rx Available? Imaging Capture Imaging with Dx of Yes Available ? data for review interest Yes Rx Available? Capture Rx Yes Capture all Rx’s and determine relevant Rx lisƟng Use the relevant Rx lisƟng to match Any relevant Rx? against Rx’s for subjects with No Dx Subject idenƟfied R135 Dx Of with the Dx of Interest Interest PopulaƟon FIGURE 10.2 Committee’s process to identify subjects with the conditions of interest. Through that process, data would be collected from the DOD military health system and from the VA health system, which would allow classification of personnel according to ICD codes of interest for this study (such as codes for TBI, PTSD, depression, suicide and suicidal ideation, and substance-use disorders), as well as other records and outcome data. That plan would have required access to actual identifiers for the individuals selected as part of the sample for further study. In effect, the ability to link the selected sampled cohorts with other databases and back to the DMDC database would have been essential for the development of the committee’s analytic database. However, all committee analyses and later report data would have used only deidentified or aggregated information. SUMMARY The US government collects numerous useful datasets. Understanding how to gain access, the requirements for receiving data, what information is in the databases, and the terminology used can be difficult, expensive, and time-consuming. Time became an issue for the committee, and it abandoned most of its data-collection efforts; the committee considers that an important missed opportunity. In the end, the committee relied primarily on the published

OCR for page 457
PROPOSED DATA ANALYSES 469 literature to respond to the legislation; however, with respect to many issues of concern, there was a paucity of published data. Limited analyses, that is, descriptive analyses of those deployed to OEF and OIF were completed. RECOMMENDATIONS There has been little quantitative characterization of the issues described in the legislation, but the committee identified a wide array of data and databases available in DOD, VA, and other federal agencies that could be used to address many of the questions posed by the legislation that motivated its work. On the basis of available data, the committee developed a comprehensive data-analysis plan. The committee notes that in addition to its recommendation for comprehensive data analyses, privacy experts will need to be involved with data owners before data are linked and made accessible to researchers. The committee believes that privacy and confidentiality are essential alongside issues of coordination and synchronization of data sources. The committee recommends that the Department of Defense and the Department of Veterans Affairs support comprehensive analyses of relevant data that reside in the two departments and other agencies of the federal government. Their databases should be linked and integrated so that they can be used effectively to address questions regarding readjustment that are not answered in the peer- reviewed literature. The committee’s preliminary work in this area has provided a clear rationale, justification, and roadmap for comprehensive data analyses. Comprehensive data analyses will require establishment of systematic, timely processes for using available government data and linking them in such a way as to improve the characterization of issues of interest. No databases or files fully integrate basic deployment and demographic data with data on health outcomes, treatment or transition-of-care files, data on access to care, records of employment before and after deployment, and data on other processes and outcomes. A comprehensive analytic database will have to be created and maintained. The committee recommends that the secretary of defense and the secretary of veterans affairs establish an interagency work group to identify and examine the feasibility of linking data that exist in Executive Branch departments and agencies throughout the federal government. The work group should be tasked to explore issues related to coordination among agencies, for example, defining common goals, establishing common policies and procedures, creating mechanisms for data-sharing, establishing records systems, and overcoming legal impediments and meeting legal requirements. The work group should provide the secretaries with options and recommendations for establishment of a sustainable program for long- term cooperation and data-sharing to improve understanding of the outcomes of military service and readjustment after combat deployment. The committee believes that many of the issues examined in this study can be addressed through analyses of data already maintained by numerous federal agencies. The committee tried

OCR for page 457
470 RETURNING HOME FROM IRAQ AND AFGHANISTAN to gain access to the data files so that it could begin such analyses, but it faced numerous obstacles in its attempts to access them. In light of those difficulties, the committee recommends the following actions to address many of the problems that it faced. The committee recommends that clear procedures be developed for accessing data held by the Department of Defense, the Department of Veterans Affairs, and other federal agencies. The procedures should appear on each agency’s website with access to its data dictionaries. That would enable researchers and others wishing to access data to understand all the requirements before they begin their data- gathering efforts and would provide information about the types of data that are available and how to access them. The questions posed to the committee are complex and critical to the well-being of US veterans, their families, and the communities in which they live. A major finding of the committee is that there is no way to provide data-based answers to those questions. All agencies that collect, store, and manage information relevant to veterans and their families should give high priority to coordination of those efforts throughout the federal statistical system so that informed decisions about veterans’ readjustment needs can be made in the near future. The committee believes that such coordination will greatly enhance the ability of researchers and the government to link data held by multiple agencies to allow the types of analyses recommended above.