Connecting Contemporary Paradigms to the Social Security Administration’s Disability Evaluation Process

Abstract

From 1998 to 2008, the Social Security Administration’s (SSA) disability insurance program (DI) applications rose from 1.2 million to 2.3 million and exceeded 3 million in 2009. Given these large and growing numbers, even small changes in processing disability applications may reduce processing time, lower program costs, and improve performance of SSA’s disability programs. A literature review examining current conceptual models of disability and SSA’s disability evaluation process for adults was conducted. A gap exists between contemporary models of disability and how SSA defines and operationalizes disability. This is complicated by substantial variation in the timing, quantity, and quality of applicant functional information and workplace demands. A focus on impairment marginalizes more comprehensive assessment of function necessary to assess capacity for work. Novel assessment methodologies, such as computer adaptive testing to measure human functioning may hold promise for SSA’s data collection methods and disability assessment.

The Social Security Administration (SSA) disability programs, Social Security Disability Insurance (DI) and Supplemental Security Income (SSI), are currently the largest federal programs related to disability (SSA, 2009b). In 2007, these programs provided basic income support to more than 12 million adults and children with disabilities (SSA, 2009a). The DI program, under Title II of the Social Security Act (the Act), provides earnings replacement insurance for more than 200 million workers, covering earnings loss in the event of a work limiting disability (SSA, 2009f, Table 4c2F; see Note 1). These individuals are considered “insured” in that they contributed to the Social Security Trust Fund through taxes on earnings. The SSI program, under Title XVI of the Act, provides financial support to low-income adults and children who are blind and/or have a disability as well as low-income adults older than 64.

DI and SSI are important sources of financial and health care support for people with disabilities. In 2007, the average monthly payment for DI and/or SSI beneficiaries ages 18 through 64 was approximately $890 (SSA, 2008a, Table 65). For a majority of beneficiaries, these payments represent more than 75% of their total income (Social Security Advisory Board [SSAB], 2006b). In addition, DI beneficiaries are eligible for Medicare health care benefits after a 24-month waiting period. Also, in a majority of states, SSI beneficiaries are immediately eligible for Medicaid (SSA, 2009f).

Over time, these programs have seen a dramatic rise in the number of applications. From 1998 to 2008, DI applications rose from 1.2 million to 2.3 million (SSA, 2009f, Table 6c7) and exceeded 3 million in 2009. Partially as a result of this growth, applicants must now wait months to years for a final decision. A significant cause of determination delay is difficulty obtaining medical records (U.S. Government Accountability Office [GAO], 2008). By regulation, the SSA must use medical evidence to confirm impairment severity. In 2007, the SSA spent $123 million on the collection of medical records alone. Although challenges in obtaining medical evidence are common, inconsistency in the evidence obtained further complicates the determination process (GAO, 2008). Given the escalating numbers, even small changes in the way that disability applications are processed may lead to reductions in processing time and program costs, improving the overall performance of the SSA disability programs.

The purpose of this article is to advance conceptual perspectives of disablement in the nation’s largest disability programs by exploring conceptual and operational conflicts stemming from the difficulty of assessing work disability within current statutory guidelines. We identify conceptual gaps in the statutory definition of disability provided by the Act, comparing it to contemporary models of disability and functioning. We consider the importance of these conceptual gaps, focusing on the sequential evaluation process, but appreciate the influence of these gaps relative to the content of evidence collected, the decision rules applied, the length of the disability determination process, and the appeal of denied disability claims. Next, we review efforts at the SSA to improve the disability determination process. Finally, we recommend novel ways to improve the process based on the reform efforts to-date, our evaluation of the process’s current limitations and new technologies available in the marketplace.

Conceptual Issues Related to Disability

In compliance with statutory guidelines, SSA frames disability as work disability. Prominent conceptual models of disability (Altman, 2001) include those proposed by Nagi (1965, 1991), Verbrugge and Jette (1994), the Institute of Medicine (IOM; Brandt & Pope, 1997; Pope, 1992), Abberley (1987), Oliver (1990, 1993, 1996), and the World Health Organization (1980, 2001). Although aspects of these models differ, there is agreement that disability cannot be viewed as an individual attribute (Zola, 1993). Rather, contemporary concepts of disability depict it as the outcome of the interaction between people with potentially limiting health conditions and their environment (Brandt & Pope, 1997). In other words, disability should be thought of as the difference between individual capabilities and environmental demands (Verbrugge & Jette, 1994). The IOM (1991) notes that disability is not a stable attribute across situations since functioning is influenced by environments (Brandt & Pope, 1997). Disability is a complex process, which is multidimensional, dynamic, biopsychosocial, and interactive in nature. It is not a dichotomous phenomenon of presence or absence, although it is often characterized and measured as such (Zola, 1993).

With the exception of the social model (Abberley, 1987; Oliver, 1990, 1993, 1996), disablement models are typically composed of biopsychosocial conceptual components that can be thought of as building blocks. Models of disablement typically encompass the interaction between a health condition (pathology) at the cellular or tissue level, impairment at the organ or system level, and functional limitation at the person level with contextual factors (societal level involving interaction with environments; Nagi, 1991; Verbrugge & Jette, 1994). Disablement and functioning are not static states but create a continuum of human functioning influenced by intraindividual and extraindividual factors (Jette & Badley, 2000; World Health Organization, 2001).

Movement along the continuum—the degree to which changes in function at one point in the continuum affect function at other points in the continuum—is influenced by personal and environmental barriers and facilitators. A reduction in function, at any point along the continuum, may be restored via rehabilitation and environmental modification. Intervening factors such as medical treatment, use of assistive devices, or the occurrence of secondary conditions can influence decline and recovery, that is, movements in either direction along the continuum (Brandt & Pope, 1997).

Statutory Definitions

The SSA statutory definition of disability differs considerably from the description of disability found above. The most significant difference is that SSA defines disability at the impairment level, equating disability with work disability. From SSA’s standpoint, disability refers to an inability to engage in any substantial gainful activity by reason of any medically determinable physical or mental impairment that can be expected to result in death or that has lasted or can be expected to last for a continuous period of not less than 12 months (SSA, 2009c; SEC. 223, 42 U.S.C. 423(d) (1)). Individuals are disabled only if their physical or mental impairment or impairments are of such severity that they not only are unable to do their previous work but also cannot, considering their age, education, and work experience, engage in any other kind of substantial gainful work that exists in the national economy, regardless of whether such work exists in the immediate area in which they live, or whether a specific job vacancy exists for them, or whether they would be hired if they applied for work (SEC. 223, 42 U.S.C. 423(d)(2)(A)). Physical or mental impairment is defined as an impairment that results from anatomical, physiological, or psychological abnormalities, which are demonstrable by medically acceptable clinical and laboratory diagnostic techniques (SEC. 223, 42 U.S.C. 423(d)(3)).

SSA’s link between impairment and disability has important implications. The Act situates disability at the level of impairment. However, contemporary models of disability recognize disablement as an interaction between the person and the environment. The implications of this conceptual gap are that the severity of anatomical, physiological, psychological abnormalities may not predict the ability to work. Saad Nagi (1965) gives the example of two individuals with loss of a finger; the loss resulted in inability to work for the concert pianist but had no adverse effect on the teacher. In other words, two individuals with the same impairment can have very different work disability. Since impairments alone may not predict the ability to work, the early steps of SSA’s sequential evaluation process may lead to false positives and award claimants with anatomical, physiological, psychological abnormalities that are able to engage in substantial gainful employment. On the other hand, a sole focus on impairments may also lead to false negatives and deny appropriate claimants because their impairments do not fit into the required categories. These cases, some of which will be reversed on appeal, might have been adjudicated faster if the SSA focused on function first rather than impairment.

Operational Definitions

SSA translates the concepts set forth in the Act into measures suitable for identifying eligibility via a five-step sequential evaluation process. Steps 1 through 4 place the burden of proving work disability on the claimant. Since SSA defines disability as a dichotomous state (yes or no), there are no partial awards (SSAB, 2003, p. 18). This section provides an overview of the sequential evaluation process and demonstrates how conceptual gaps may be embedded and perpetuated within the process itself. A summary of the five-step process and decision rules can be found in Table 1.

Table 1.

The Social Security Administration’s Five-Step Sequential Evaluation Process

Decision sequence	Decision outcome	Decision outcome
1. Substantial gainful activity	No—proceed to Step 2	Yes—not disabled
2. Severe impairment	Yes—proceed to Step 3	No—not disabled
3. Meets or equals listing	No—proceed to Step 4	Yes—determination of disability
Residual Functional Capacity assessment
4. Engage in past work	No—proceed to Step 5	Yes—not disabled
5. Engage in any work	No—determination of disability	Yes—not disabled

Filing a Claim

Claimant application forms request information about the claimant’s illnesses, injuries and/or conditions (dates and impact on ability to work), medical providers and medical conditions, treatments and tests, work history (job title, industry, time period, hours per day, days per week, pay, and job requirements), education or training, and vocational rehabilitation or employment support services. An optional Function Report (SSA-3373-BK form) may be filed by the claimant, although this form is not included in the Starter Kit or other online resources.

Advocacy organizations and legal advisors recommend inclusion of a completed Function Report with the claimant application. This form asks about difficulty performing daily activities, such as personal care, meal preparation, yard work, and social activities; basic abilities, such as lifting, squatting and bending, getting along with others, coping with stress, and memory; and the use of devices (crutches, walker, wheelchair, cane, brace or splint, artificial limb, hearing aid, glasses or contact lenses, artificial voice box, and other devices). Although the form captures functional information, it does not link functional abilities with the applicant’s specific occupational demands.

Step 1: Substantial Gainful Activity?

The first step of the sequential evaluation occurs at the SSA Field Office, where SSA staff ascertains whether an individual meets the basic requirements of the program(s) based on citizenry, financial assets (SSA, 2008b), and absence of substantial gainful employment. If these requirements are met, SSA sends the claim to the Disability Determination Services (DDS) office in the claimant’s state.

Step 2: Determine the Impairment

In the second step, the DDS documents the presence of an impairment or combination of impairments severe enough to interfere with work activities (SSA, 2009g). In addition to information from the initial application forms, DDS officials may collect additional medical records from the claimant’s doctors, hospitals, clinics, or institutions where the claimant has been treated. The DDS may also request a claimant’s health care provider indicate the applicant’s primary medical condition(s); when the condition(s) began; how the medical condition(s) limits the claimant’s activities, including work-related activities, such as walking, sitting lifting, carrying, and remembering instructions; results of related medical tests; and what treatment the claimant has received. Medical evidence must demonstrate an association with the claimant’s medical condition (SSA, 2007b).

Determination of impairment must be made by “acceptable medical sources” (SSA, 2008c). Acceptable sources include licensed medical professionals such as physicians, osteopaths, psychologists, optometrists, podiatrists, and speech-language pathologists (SSA, 2009d). Licensed clinicians with particular expertise in assessing impairment and resultant functional implications, such as physical and occupational therapists are not included in this list. Although SSA may use evidence from experts in function, like physical and occupational therapists, once a diagnosis has been obtained from an “acceptable” medical source (SSA, 2009e), this critical information may be missing unless it is actively sought out.

Should a claim be advanced to the second step of the evaluation process, yet lack sufficient evidence from the claimant’s own medical sources, the DDS may request a consultative examination (CE) with a medical provider. CE exams are used to augment existing evidence if the existing evidence is inconsistent or unproductive or if the claimant can justify the need for another opinion (SSA, 2008c, p. 15). Overall, a CE focuses on detecting impairment and broad changes in functional activity but does not necessarily link these deficits to workplace demands and associated functional capacity for work. Although a CE may be completed by the same medically accepted sources as noted above, SSA gives priority to evidence from a claimants’ treating physician. The objective of the CE is to inform subjective claimant information with clinical findings. CE guidelines for the assessment of functional performance vary by medical discipline (SSA, 2009d). Overall, CEs focus on detecting impairment and broad changes in functional activity but do not link these deficits to workplace (environmental) demands and associated functional capacity for work.

Step 3: Does the Applicant Meet or Exceed the Listings of Impairments?

The third step of the sequential evaluation process assesses the claimant’s impairment relative to SSA’s Listing of Impairments (aka the Listings, medical listings) that will qualify an applicant for disability. The Listings describe diagnoses and impairments associated with major body systems considered severe enough to limit work activity (SSA, 2008c, p. 19). For example, criteria for Listing 1.07, fracture of an upper extremity, specify that a nonunion fracture under continued surgical management without restoration of function and with the expectation that function will not be restored within 12 months of onset must be met for award (SSA, 2008c, p. 34). The Listings are used only in the third step of the sequential evaluation process (SSA, 2008c, p. 19).

The claimant’s impairment must be found to meet the criteria established in the Listings or deemed equal in severity. For an impairment to equal a Listing, the signs, symptoms, and laboratory results of the alleged impairment must be equivalent with those of an analogous Listing (SSA, 2009g). In addition, a combination of impairments may be determined to “equal” the Listings. In this circumstance, the signs, symptoms, and laboratory findings of the alleged impairments must be matched to closely related Listings. The validity of awards determined by meeting the Listings versus equaling the Listings requires further investigation.

Recent evidence suggests the diagnostic basis for the Listings has become less useful over time to SSA. Data from the early years of the program indicate the Listings accounted for more than 90% of allowances. By 2004, use of the Listings leading to awards dropped to 52% (Stobo, McGeary, & Barnes, 2007, p. 4). A 2007 IOM report notes substantial award variation among conditions for claims determined according to the Listings and recommends the Listings be validated against the ability to sustain gainful employment (Stobo et al., 2007, p. 55).

If a claimant’s application is not found to meet or equal the Listings, a Residual Functional Capacity (RFC) assessment is conducted prior to progressing to the fourth step of the sequential evaluation. The RFC assesses the impact of impairment on functional capacity to work. The RFC itself does not consider potential mediating conditions such as age, gender, baseline levels of conditioning, and the actual or potential use of assistive devices or environmental modifications. Although the CEs used in Step 2 may assess functional activity such as gait, bending, and writing to discern severity of impairment, the RFC assesses similar functional ability relative to work capacity (Hu, Lahiri, Vaughan, & Wixon, 2001, p. 349).

The RFC is completed by a medical or psychological consultant, or an SSA disability examiner in some states. Functional capacity is determined according to medical evidence addressing signs, symptoms, and laboratory findings (SSA, 2007b, p. 57). Claimant allegation and field officer observation may be considered, particularly if consistent with the medical evidence. There are two types of RFC assessments that examine claimant capacity for work. A physical assessment examines exertional limitations (considered strength activities such as walking and lifting) and nonexertional limitations (physical movements such as reaching and mental functions such as remembering instructions and social interaction). The physical RFC is completed by a medical consultant and must specify each of the claimant’s functional limitations relative to capacity for work (SSA, 2007b, p. 62). The mental RFC must be completed by a psychiatrist or psychologist and documents limitations in mental functioning important for the work environment. Mental functioning is assessed relative to the demands of unskilled, semiskilled, and skilled work environments.

SSA uses the RFC to assess claimant functioning relative to work capacity in Steps 4 and 5 of the sequential evaluation process. For example, a claimant is said to possess a medium work capacity if he or she is able to occasionally lift 20 to 50 pounds, can frequently lift or carry 10 to 25 pounds, can frequently stoop or crouch, and can tolerate standing and walking 6 out of 8 hours. A claimant possessing medium work capacity is considered suitable for sedentary and light work as well. Basic mental abilities for any work include the ability to carry out simple instructions, make simple decisions, ask simple questions, and get along with coworkers. The RFC focuses on work-related limitations as a result of a medically determined impairment or impairments (SSA, 2007b, p. 76). The objective of the RFC is to determine the claimant’s highest level of work capacity.

Step 4: Past Work?

Step 4 compares the claimant’s past relevant work (PRW) with outcomes of the RFC through two lenses: the ability to conduct work activity in a manner consistent with past performance and the ability to conduct work activity according to acceptable standards in the national economy. At this step, a claimant’s demographic characteristics such as age and education are not considered (Johns, 2009). If a claimant is found capable of PRW, benefits are not awarded. If the claimant is unable to complete past work, the claim progresses to Step 5 of the sequential evaluation process.

Step 5: Any Work?

The fifth step of the sequential evaluation process is to determine the claimant’s ability to engage in any work. At this step, the burden of proof falls on SSA to prove the claimant can engage in substantial gainful activity available in the national economy. Claimant characteristics such as age, education level, and transferable skills are viewed as relevant considerations in the ability to engage in a new work activity.

Substantial gainful employment is generally defined as work that involves significant physical and/or mental activities and is performed for pay averaging more than $980 a month (for 2009) for impairments other than blindness (SSA, 2009i). The Act acknowledges the existence and relevance of an occupational environment bounded by the types of jobs that exist in the national economy in which people with disabilities (and those without) must operate. Steps 4 and 5 of the sequential evaluation process seek to comply with statutory guidelines by linking RFC to employability. However, the way in which the environment of the work place is integrated into the disability evaluation process and used for decision making is complex.

Historically, SSA has relied on the Dictionary of Occupational Titles (DOT) of the U.S. Department of Labor (DOL) to link functional ability to the work environment. The DOT was established to assess occupations in the national economy and provide a description of worker characteristics and environmental conditions associated with those occupations (Miller, Treiman, Cain, & Roos, 1980). The DOT has not been substantively revised since 1977 (Karman, 2009). Since that time, occupations in the United States have shifted from industrial to technological, with a concomitant shift in functional and educational requirements. Data from 2007 indicate that in 50% of initial claims, the DOT is inadequate to resolve issues and a subsequent vocational review is required (Karman, 2009). Frequent litigation related to this portion of the Act (Blair, 2009) also testifies to the agency’s difficulty aligning features of the environment with medical evidence.

More recently, the DOL is replacing the DOT with O*NET, a substantially revised and electronic version of the DOT. However, since neither the DOT nor O*NET were created for the purposes of disability evaluation (Pfaff, 2009) the SSA is still challenged with establishing mechanisms to assess claimant RFC relative to a dynamic work environment. In this regard, SSA has convened internal and external agency experts forming the Occupational Information Development Advisory Panel. The Panel will formulate recommendations for an occupational information system better suited to meet the needs of SSA’s disability programs.

The five-step sequential evaluation process is used to make an initial disability determination. If a claimant disagrees with the determination outcomes, there are four levels of appeal. A request for reconsideration may be submitted, whereby the case and any new evidence will be reviewed by another DDS team. Should a claimant disagree with outcomes of the reconsideration, a hearing before an administrative law judge (ALJ) may be requested. If a claimant is dissatisfied with the decision of the ALJ, the claimant may proceed to the next step in SSA’s appeals process, a review by the SSA Appeal’s Council. With continued dissatisfaction, the claimant may file a civil suit in federal district court (SSA, 2008e).

The sequential evaluation and appeals processes are interactive in nature; therefore, a change to the sequential evaluation process may influence the appeals process as well. Over time, SSA has recognized the need for programmatic change and has put forth substantial effort to enhance efficiency of its disability programs. These efforts vary in scope and complexity, influencing claimants as well as SSA policies, procedures, and resources.

SSA’s Redesign Efforts

Recognizing that the disability determination process needs improvement, the SSA has undertaken a number of significant internal and external reviews. This section provides an overview of efforts influencing the sequential evaluation process, with particular focus on how SSA attempted to improve information collected in the DDS process.

The 1994 Redesign

An initial redesign effort occurred in 1993, when Commissioner Shirley Chater launched the Disability Redesign project. Under this effort, SSA formed a Disability Re-engineering Team, which examined how the agency could improve the quality, timeliness, and efficiency of the determination process. This team issued a report in September 2004 that called for 83 initiatives to be accomplished over 6 years, with 38 to be implemented within the first 2 years (SSA, 2000, p. 7; SSA, 2002; SSAB, 1998). The 1994 effort is noteworthy both because of the comprehensive nature of its scope and because subsequent SSA redesign efforts either extended or adapted the initiatives originally introduced in 1994.

The report called for SSA to redesign many aspects of the disability determination process except the statutory definition of disability, benefit amounts, the use of ALJs, and vocational rehabilitation. Most importantly, it brought the measurement of applicant function to the forefront, proposing a New Disability Decision Methodology. This new process was intended to replace the Listing of Impairments with a new and standardized measure (or set of measures) of a claimant’s ability to function (SSAB, 1998, p. 52). Between 1994 and 1998, the agency continued to test many of these initiatives but with limited plans to implement them (SSA, 2002, pp. i–ii; SSAB, 1998, p. 49). Although the importance of functional assessment emerged in redesign efforts, mechanisms to capture information in an efficient manner without administration by specialists were not identified (Wunderlich & Rice, 1998). These issues can now be addressed with the use of new technologies, such as computer adaptive testing (CAT) tools.

From an information flow perspective, the 1994 redesign sought to bring all relevant information to the earliest point in the evaluation process. By introducing the Disability Claims Manager (DCM), who handled both nonmedical and medical aspects of claims at the initial application stage, the agency attempted to consolidate decision making to a single point. In addition, by proposing that the Listings be replaced by standardized metrics that capture individual functional capacity, the agency made an effort to obtain these data at the earliest stage of the evaluation process. Although the 1994 redesign effort was clearly an ambitious endeavor, proposed changes to the determination process did not fully address the need to link function with workplace demand.

In 1996, GAO conducted a thorough review of SSA’s 1994 redesign efforts and concluded that the complexity of the proposed initiatives limited their utility (GAO, 1996). SSA’s own Office of the Inspector General (SSA-OIG) also recognized that the agency had some difficulty in obtaining support from its key stakeholders. SSA-OIG reported that as a consequence, the agency had to scale back its original initiatives and proposed a smaller set to be implemented over the following 9 years (SSA, 2002, pp. 1–2).

The 1999–2000 Redesign

In March 1999, Commissioner Ken Apfel announced a plan to implement some of the 1994 plans by creating ten Prototype DDSs, as well as Quality Assurance (QA), DCM, Process Unification programs, and a Hearings Process Improvement processes, with plans to phase in these implementations throughout FY2000 and into FY2001 (SSA, 2002).

The 2001–2003 Redesign

In 2001, SSAB issued a special report calling for additional in-depth reviews as well as fundamental reforms of the disability determination system. The report focused specifically in the areas of the quality and consistency of decisions as well as the medical basis of decisions. It reported that between 1983 and 2000, the percentage of DI cases determined by function dramatically doubled, noting a “persistent trend away from decisions based on the Listing of Impairments to decisions that increasingly involve assessment of function” (SSAB, 2001).

In the same year, Commissioner Apfel launched another agency-wide effort to reform the process, by enhancing the DDS Prototypes and introducing changes such as (a) allowing the DDS disability examiner to make the initial determination of disability without requiring the certification of a medical consultant, (b) introducing a “Claimant Conference” during which a claimant may interact with the disability decision maker at the start of the process, (c) enhancing documentation and explanations by the DDS, and (d) eliminating a Reconsideration Step (SSAB, 2001).

By 2002, the agency reported that more allowances were made at the 10 prototype DDSs, but productivity decreased by 13%, processing time increased by 23%, and appeals increased by 6%. SSA-OIG also reported higher than average attrition rate among Disability Examiners at 60% in prototype DDSs (SSA, 2002, p. B-8). Since no baseline comparison or randomized control was established between the 10 prototype DDSs and their regular counterparts, it is difficult to know if these changes were the result of the redesigned process, some systematic differences, or simply chance (SSA, 2002, pp. 14–16). Nevertheless, SSA-OIG recommended that SSA proceed to national implementation, and SSA concurred (SSA, 2002, p. 19).

The 2004–2006 Redesign

In 2003, SSA launched an initiative to implement an efficient paperless disability claims processing system known as eDib. Under this effort, the large volume of documents traditionally stored in paper folders were to be stored, viewed, and processed electronically, in the hope that doing so would improve the speed of the adjudication process (SSA, 2006b, p. 1). The new approach was named the Disability Service Improvement (DSI) initiative (SSAB, 2006a).

In August 2006, SSA began a phased implementation of DSI. In conjunction with implementation, the agency also initiated a comprehensive evaluation of the new process. This evaluation was intended to, among other things, (a) determine whether DSI can meet the goals SSA has established for improving claims processing timeliness, accuracy, consistency, and productivity; and (b) assess other effects of DSI including its impact on administrative costs, applicant satisfaction, employee satisfaction, and DI and SSI caseload characteristics (SSA, 2006a, p. 1).

The 2007–2008 Redesign

In July 2007, SSA’s DSI Initiative was discontinued on a national level and a new effort called Intelligent Disability was launched, with focus on four major areas of reform, namely (a) compassionate allowances, (b) improved hearing procedures, (c) increased adjudicatory capacity, and (d) improved efficiencies with automation and business process (SSA, 2007c). The objective of the new effort was to ease system pressure on adjudicative review, specifically the hearings level, and enhance electronic processing of claims.

The IOM Report

Recently, the SSA asked the IOM to make recommendations for improving the timeliness and accuracy of its decisions. The 2007 IOM report titled “Improving the Social Security Disability Decision Process” reflects the findings of this committee and its recommendations (Stobo et al., 2007). The IOM committee noted that, as medical treatment and assistive technologies advance, the diagnostic basis for the Listing of Impairments has become less useful a marker of disability. This finding led to the committee’s first recommendation that the performance of the Listings be examined for their validity in expediting awards and as a measure of work disability.

The committee recommended that SSA support the development of promising alternative approaches, including the creation of new screening tools as well as the utilization of a generic functional evaluation as it relates to a claimant’s ability to engage in substantial gainful activity. As part of SSA’s sequential evaluation process, allowance can be justified by the presence of a combination of impairments equal in severity to a listed impairment. The committee suggested that the most practical way of deciding such cases would be to assess the net functional impact that the impairments have and recommended functional assessment of these applicants. Finally, the committee recommended that SSA monitor and sponsor research regarding the extent to which medical and functional criteria are correlated with limitations in performance of substantial gainful employment.

As the number of disability claims continues to escalate, continued effort to improve SSA’s disability determination process is crucial to meet demand. The processes by which work disability are determined require continued scrutiny as exogenous conditions change over time. Redesign efforts must address the early stages of the sequential evaluation process to ease system pressures that mount with progression through the process (SSA, 2008d, p. 4).

Consequences to Applicants

Through the years, SSA’s reform efforts demonstrate the agency’s commitment to establishing smooth and equitable processes. However, SSA’s efforts have been hampered by the enormity of the constituent pool and the complexity of translating statutory language into an efficient determination process. As a result, the application process is still unduly long and does not include functional status information in a standardized or timely fashion. There are three related concerns about the disability determination sequential evaluation process that are potentially influenced by the way in which functional information is collected: (a) the expense, variation, and timeliness of collecting materials, (b) the timeliness of decisions, and (c) the accuracy of those decisions.

In FY2007, SSA reimbursed states more than $400 million for expenses associated with obtaining medical records and consultative exams (GAO, 2008, p. 2). In a survey of DDS directors, the GAO (2008) reported that 14 of the 51 DDSs reported not receiving records for at least 20% of requests in FY2007. In addition, 41 of 51 DDSs reported routinely asking claimants’ own medical providers to perform consultative exams; yet 34 reported that medical providers never or almost never agree to do so. With regard to processing time, SSA’s 2009 targeted average processing time is 103 days for initial claims, 506 days for ALJ hearings, and an additional 242 days for appeals of ALJ decisions (SSA, 2009h). These targets are based on SSA’s budget and anticipated case loads. The SSA is actively developing processes to retrieve, store, and access medical records electronically (Feldman & Horan, 2010; GAO, 2008). This advance is intended to enhance efficiency of record submission by providers as well as reduce processing time of the claim. The SSA is presently storing medical records for new claims in electronic format, permitting remote retrieval and review that may enhance the consistency in policy interpretation (Feldman & Horan, 2010). Although electronic records may substantially affect efficiency and consistency of the disability evaluation process, it does not influence the quality of the information obtained.

Relative to accuracy, the decision to allow or deny disability benefits has four possible outcomes: the allowance of an eligible claimant (true positive), the denial of an ineligible claimant (true negative), the allowance of an ineligible claimant (false positive), and the denial of an eligible claimant (false negative). False positive and false negatives represent inaccurate decisions. Each year SSA reviews a sample of initial determination cases for QA purposes. From this review, SSA calculates the net accuracy rate, which is the number of reviewed cases in which the allowance or denial was overturned as a percentage of all reviewed cases. SSA targeted net accuracy rate for 2009 is 97% of reviewed cases. It is important to note that although 97% accuracy may appear to be good, with 3 million applicants a year, 30,000 cases will still be a false positive or false negative.

However, the subsequent allowance of an original denial may also be viewed as a signal of inaccuracy and will certainly affect the issue of timeliness. Unfortunately, there is no gold standard of disability measurement to judge whether allowances on appeal represents a false negative (because of a prior inaccurate denial) or a false positive (inaccurate award on appeal). At best, an allowance on appeal suggests that the procedures are different and/or uniform rules are being applied differently, assuming no additional information is collected and/or there are no changes in a claimant’s condition.

Nevertheless, reversals on appeals are viewed as problematic. These appeals can be quite lengthy, requiring the applicant to wait in a “nonworking” mode for months to years. In addition they are costly. Those appeals going to the ALJ often require legal counsel (paid for by SSA in SSI claims), and counsel often retains a large portion of any back payments, denying the applicant his or her full insurance payment.

Figure 1 characterizes the progression of allowances, denials, and appeals. Among DI disabled workers and SSI-only aged 18 to 64 claims filed in the calendar year 1999, 989,413 claims were forwarded to the DDS system for determination (SSA, 2007d, Tables 69–71; SSA, 2007a, Tables 60–62). Of these claims, 30% were initially allowed and 48% of the denied claims were the subject of a reconsideration request. Of claims reconsidered, 14% were allowed. Of the reconsiderations denied, 73% were the subject of an appeal to the ALJ level and above. Furthermore, 61% of these appeals were allowed. The role of additional information and/or changes in claimant status influencing higher allowance rates further into the appeals process must be considered.

Figure 1.

Medical decisions among DI disabled workers and SSI-only aged 18-64 claims filing in 1999

Summary

As we have seen, the sheer number of SSA applicants contributes to long wait times and inappropriate determinations. However, there are also conceptual gaps and operational issues in the process that may exacerbate the problems experienced by applicants. These are listed below.

1. There is a gap between contemporary notions of disability and how SSA defines and operationalizes disability. SSA’s statutory definition of disability is based on physical or mental impairments caused by health conditions, whereas contemporary models depict disability as the gap between an individual’s functional abilities and his or her environmental demands.

2. There is significant variability in the timing, quantity, and quality of the functional information obtained about the applicants.

3. There is significant variability in the timing, quantity, and quality of the information gathered about an applicant’s environmental (workplace) demands.

4. Information from care providers such as physical and occupational therapists, who may have expertise in functional assessment, is marginalized as SSA focuses on identifying impairments.

5. There is an inappropriate focus on a single impairment as the primary reason for disability with inadequate methodologies to account for the functional consequences of secondary conditions, multiple comorbidities, or multiple impairments.

6. The DOT and O*NET may not accurately reflect current workplace functional requirements.

7. The Listing of Impairments should be expanded to include more comprehensive functional criteria.

Recommendations

Ideally, the SSA would consider altering their statutory definition of disability to reflect more contemporary concepts. Defining disability based on an applicant’s functional abilities and environmental demands could shorten the process for many, reduce the number of inappropriate determinations, and ultimately give the SSA a stronger foundation from which to grow. However, given the legislation required to make this change, it is unlikely that this will occur in the near future.

Although SSA is unlikely to change its definition of disability, they can still alter the manner in which an applicant’s functional status is assessed and the timing used to interpret these data. Although SSA currently strives to collect information about an applicants’ functional abilities, the Disability Report is optional and the RFC is done late in the process and does not address all relevant domains of functioning. Thus, SSA has, at best, an incomplete picture of an applicant’s functional status as it relates to his or her workplace. To complete this picture, SSA must consider obtaining additional information from applicants.

Collecting additional functional information about all SSA applicants may seem to be a daunting challenge. However, using novel assessment methodologies available in the marketplace, the additional data may actually shorten the process for many applicants and, as they are computer driven and available in real time, could substantially relieve overburdened staff. One such promising approach is item response theory (IRT) using CAT. The efficient collection of functional information provides a mechanism to enhance the quality of the data collected and the accuracy of decision outcomes.

IRT-CAT Technology

IRT-CAT is a method that is widely used in academic settings and has recently been applied to the measurement of human function (Cook, Monahan, & McHorney, 2003; Hambleton, 2000; Hays, Morales, & Reise, 2000; Revicki & Cella, 1997; Sands, Gade, & Knapp, 1997; Wainer et al., 2000). Since there is a broad range of function to be measured relative to work capacity, it would be hard to design a set of traditional instruments that included the number of items necessary to precisely measure all applicants. The resulting length and complexity of any such battery would raise concerns over respondent burden and administration costs. IRT-CAT represents simple artificial intelligence software that reduces these concerns.

An IRT-CAT program begins with an initial question or item. The software then tailors the subsequent assessment by asking only the most informative questions based on a person’s response to previous questions. For individual scales, scores can be computed in a few minutes with a few items, considerably reducing respondent burden. Each individual is scored on the same underlying outcome continuum so that results can be compared to other individuals or the same individual assessed over time. Results can be displayed instantly for immediate interpretation and use. The demonstrated advantages of the CAT instruments are reduced respondent burden, increased score precision, elimination of ceiling and floor effects, monitoring of data quality in real time, and lower data collection costs. These methods could also be used to collect functional assessment data from health care providers as well as applicants.

CAT assessment tools have demonstrated superior performance across domains of health relative to traditional assessment methods. A study validating a CAT tool for lower extremity osteoarthritis found the CAT instrument performed better than the current standard, the Western Ontario and McMaster Universities Osteoarthritis Index, demonstrating improved reliability, accuracy, and precision relative to functioning and pain (Jette et al., 2009). As a delivery mechanism, CAT instruments tailor items presented to each respondent, remarkably enhancing efficiency and scoring precision. In addition, the ability to assess bias of the instrument items (differential item functioning) enhances instrument validity by comparing like responses in terms of functioning within a health domain relative to respondents with varied demographic characteristics, such as race and gender (Choi, Reise, Pilkonis, Hays, & Cella, 2010). In other words, the CAT tool appears to be more sensitive to true variations in functioning compared to static form measures.

Static short form instruments for depression were found to perform nearly as well as CAT tools, although the CAT instruments exhibited improved precision at the high and low ends of functioning (Choi et al., 2010). The extensive range of items encompassed in each CAT tool minimizes the presence of floor and ceiling effects. Traditional instruments measuring health status by assessing functioning and psychosocial well-being lack a breadth of variation that may exist across a single health domain (Revicki & Cella, 1997). The improved measurement properties of IRT-based CAT instruments are a reflection of technological advances permitting the inclusion of more comprehensive domain gradations.

With any assessment tool there is a potential for adverse incentives. Although CAT instruments cannot eliminate intentional response exaggeration, the instrument can detect inconsistent responses. In addition, the completion of CAT instruments by a claimant and his or her medical provider permits the use of statistical procedures to compare and detect inconsistent CAT outcomes. The presence of inconsistent outcomes may then warrant additional testing with traditional instruments to augment information provided by CAT. CAT technology provides an efficient mechanism to detect questionable outcomes and serves as a valuable resource that may be used to further inform interpretation of application data.

The utilization of IRT-CAT technology may hold great promise to improve SSA’s current data collection methods and assessment of disability. This could potentially allow SSA to collect more relevant and precise data in a faster, more efficient fashion. It would likely improve the uniformity of decisions by standardizing test administration and work as a mechanism to bridge disciplinary and geographic boundaries. Finally, it could result in reduction in costs if more accurate decisions can be achieved earlier in the evaluation process.

Work is the product of the whole person interacting in a workplace environment; hence, the assessment of function provides a mechanism to integrate contemporary perspectives of disablement. Although countries around the world are confronted with the complexities of disability determination, each system possessing unique administrative characteristics, conceptual agreement may be found among many nations. The Organization for Economic Cooperation and Development (2003) and its 30 member countries recognize disability as the interaction between the individual and the environment and recommend evaluation in that regard. In 2004, member countries of the European Union of Medicine and Assurance in Social Security convened to discuss the integration of the ICF (International Classification of Functioning, Disability and Health) into national social security programs by identifying key domains of function for inclusion in disability assessments by all participating nations (Brage, Donceel, & Falez, 2008).

Medical provider documentation serves as the evidentiary cornerstone for the SSA’s disability decision making. The newest edition of the American Medical Association Guides to Impairment Ratings currently highlights the role of functioning by integrating ICF concepts and terminology into the guidelines and recommending the assessment of functioning to inform impairment ratings (Rondinelli, 2009). The author notes, however, the lack of a brief yet comprehensive functional measure. The use of IRT-CAT tools provides a methodologically rigorous and efficient means to capture functional information in a manner that informs the SSA’s determination process while appreciating increasing demands on the SSA programs.

The SSA is challenged with determining claims according to the statutory definition of disability, the inability to perform substantial gainful activity. A committee formulated to review the SSA’s disability evaluation process indicates impairment alone is insufficient to meet the SSA’s objective and recommends consideration for functional ability (Wunderlich, 1999). This is particularly relevant in regard to the escalating number of claimants with mental illness. These are younger claimants whose condition is manifested by deficits in functional work capacity (Wunderlich, 1999, p. 40). In 2001, the SSA instituted the functional assessment of impairment severity as an independent criterion for award determination to child applicants (Federal Register, 2000). Applying a similar approach to the adult determination process better aligns with contemporary perspectives of disablement. Although SSA’s statutory definition of disability relies on the concept of impairment, the sequential evaluation process must conceptually link the functional consequences of impairment to work capacity (Chater, 1994; Matheson, Kane, & Rodbard, 2001). To enhance SSA’s administration of their disability programs and to better align operational frameworks, we suggest a shift in approach, where the evaluation process is reenvisioned to reflect contemporary concepts of disability. This will entail a richer characterization of applicants’ functional deficits as they relate to the workplace and inform decision making earlier in the sequential evaluation process. Functional and environmental perspectives need not exist as parallel frameworks but may be integrated as complementary mechanisms in response to diverse claimant needs and fluctuating environmental conditions.

Biographies

Diane E. Brandt, PT, MS, MA/PhD (candidate), is a protocol manager at the National Institutes of Health Clinical Research Center. Her research interests include social welfare, health policy, the implications for vulnerable populations, and the geospatial implications of policy outcomes.

Andrew J. Houtenville, PhD, is an associate professor of economics at the University of New Hampshire and the research director of the UNH Institute on Disability. His current interests include the measurement of disability and the social participation of people with disabilities.

Minh T. Huynh, PhD, is a labor economist and statistician at the National Institutes of Health Clinical Research Center. His current interests include applied statistical and econometric methods in medical research.

Leighton Chan, MD, MPH, is the chief of rehabilitation medicine at the National Institutes of Health Clinical Research Center. His research interests include health services research, health care policy, quality of care given to Medicare beneficiaries, and Medicare payment policy issues.

Elizabeth K. Rasch, PT, PhD, is a staff scientist and chief of the Epidemiology and Biostatistics Section in the Rehabilitation Medicine Department at the National Institutes of Health Clinical Research Center. Her research is directed toward promoting the health, participation, and full inclusion of people with disabilities in family and community life by informing and affecting health services, programs, and policies.