In this article, we report on the systematic review of 14 occupational therapy rehabilitation research articles published in the American Journal of Occupational Therapy (AJOT) between January 2008 and September 2009 in the practice areas of work and industry and rehabilitation, disability, and participation (Table 1). A separate review (Gillen, 2010) published in AJOT addressed rehabilitation research studies in the areas of stroke and traumatic brain injury.

To evaluate how well studies published in AJOT meet the Centennial Vision for rehabilitation research, we read and reviewed the articles and classified them by level of evidence and types of research as previously described by Gutman (2008, Table 2): systematic or narrative reviews, effectiveness studies, efficacy studies, basic research about disability, instrument development and testing, and studies with a link between occupational engagement and health.

We also summarized studies by methodology, sampling design, analytic methods, and findings (see Table 2). Levels of evidence in Tables 1 and 2 are based on AOTA’s levels of evidence rating system (Lieberman & Scheer, 2002):

• Level I: Systematic reviews, meta-analyses, randomized controlled trials

• Level II: Two groups, nonrandomized studies (e.g., cohort, case-control)

• Level III: One group, nonrandomized (e.g., before and after, pretest and posttest)

• Level IV: Descriptive studies that include analysis of outcomes (single-subject design, case series)

• Level V: Case reports and expert opinion that include narrative literature reviews and consensus statements.

In addition, we further classified the studies according to our interpretation of the priority areas identified by the Ad Hoc Work Group on Rehabilitation, Disability, and Participation (see Table 3). These categories included use of activity- and participation-level measures, use of activity- and participation-level interventions, context-based interventions, and environmental modifications.

Of the three effectiveness studies, none could be classified as Level 1 systematic reviews or randomized controlled trials. Two studies could be classified as Level 2 nonrandomized studies. One study could be classified as a Level 3 nonrandomized, one-group pretest–posttest design. No studies were Level 4 descriptive studies (see Table 1).

Fourteen articles included in this review were classified into the following categories: 5 instrument development and testing studies, 3 effectiveness studies, and 5 basic research studies. None could be classified as an efficacy study or systematic review. One study established a link between occupational engagement and health. One study was qualitative, and the remaining 13 were quantitative (see Table 2).

Only half of the studies could be classified according to the priority areas established by the AOTA working group (see Table 3). Six studies used activity- or participation-level measures, 4 incorporated activity- or participation-level interventions, 4 used context-based interventions, and only 2 included environmental modifications. Seven of the 14 studies did not fall into any of these areas.

In recent years, most studies in this area of rehabilitation have tended to focus on instrument development and validation. In addition, the instrument development studies typically used small (between 30 and 130) convenience samples from a single setting or geographic region that are not necessarily representative of the patient groups they address. The three intervention studies used quasi-experimental and cohort designs, in part, their authors noted, because of the challenge of randomizing in real-world community and employment contexts (Darragh, Harrison, & Kenny, 2008; Gentry, 2008; Kielhofner, Braveman, Fogg, & Levin, 2008). Occupational therapy has a long tradition in qualitative methods, so the limited number of studies using this methodology is somewhat surprising. By contrast, occupational therapy has had less of a tradition of population-level studies, so it is encouraging to see epidemiologic methods such as those used in Finlayson, Garcia, and Cho (2008)  beginning to be reported in the literature.

Few of the studies used strong methodological designs such as randomized controlled trials (RCTs). RCTs allow for a level of explanation of causal agents that other designs do not and often reflect the culmination of knowledge gained from numerous earlier, more exploratory studies. As such, RCTs can reflect a level of maturity and organization in thinking about a particular area of inquiry that demonstrates that a field is systematically building its knowledge base. Of the 14 studies, 3 were treatment effectiveness studies (Darragh et al., 2008; Gentry, 2008; Kielhofner et al., 2008). All 3 included activity- and participation-level interventions, and all 3 were context based. Although not a randomized controlled trial, Gentry’s (2008)  study embodied how occupational therapy uses knowledge of underlying capacity to provide context-based, participation-level interventions that improve occupational performance. Participants with multiple sclerosis were trained to use personal digital assistants to compensate for cognitive impairments and to enhance participation in everyday life tasks at home and in the community. In more rigorous research studies that included control groups, Kielhofner et al. (2008)  and Darragh et al. (2008)  provided interventions to increase productivity in community settings, supported living facilities, and the workplace, respectively. These kinds of studies, although not RCTs, exemplify the Centennial Vision for occupational therapy research by promoting participation in real-world environments.

In other studies, the use of occupation was more tangential. For example, May-Lisowski & King (2008)  had able-bodied people pick up brownie pieces to analyze upper-extremity movement. The rationale for using this food was not described. It was not clear that substituting brownies for apples or even a foam block would have made any meaningful or substantive difference to the interpretation of the study.

Of the 14 studies, 6 included activity- or participation-level assessments as the primary focus or outcome measure. The assessments included self-reported self-care and instrumental activities of daily living limitations (Finlayson et al., 2008); off- and on-road driving assessments (Kay, Bundy, & Clemson, 2008); the Laboratory Assessment Checklist (Occupational Safety and Health Administration, 2001), used in Darragh et al. (2008); upper-extremity function during everyday activities, using the Manual Ability Measure (Chen, Kasven, Karpatkin, & Sylvester, 2007) and the Test d’Evaluation de la Performance des Membres Supérieurs des Personnes Âgées (Desrosiers, Herbert, Dutil, & Bravo, 1993), both used in Rallon and Chen (2008); the Parent Disability Index (Katz, Pasch, & Wong, 2003), used in Poole, Willer, and Mendelson (2009); the Canadian Occupational Performance Measure (COPM; Dedding, Cardol, Eyssen, Dekker, & Beelen, 2004) and the Craig Handicap Assessment and Reporting Technique (CHART; Whiteneck, Charlifue, Gerhart, Overholser, & Richardson, 1992), both used in Gentry (2008); and participation in work, school, training, or volunteering (Kielhofner et al., 2008).

Three of the studies used a combination of both impairment-level and activity- or participation-level measures. For example, Gentry (2008)  used the Rivermead Behavioral Memory Test–Extended (Wilson, Cockburn, Baddeley, & Hiorns, 1989) to evaluate memory and the COPM and the CHART–Revised to capture activity and participation levels. Kielhofner et al. (2008)  evaluated both symptoms with the Revised Sign and Symptom Checklist for people with HIV/AIDS (Holzemer, Hudson, Kirksey, Hamilton, & Bakken, 2001) and productive participation with self-reports of employment status, school and training, volunteer involvement, or all of these. Poole, Willer, & Mendelson (2009)  captured impairment-level pain with a visual analog scale and fatigue with the Multidimensional Assessment of Fatigue Scale (Tack, 1991), while capturing activity or participation level with the Health Assessment Questionnaire (Fries, Spitz, Kraines, & Holman, 1980) and a parenting tool, the Parent Disability Index.

Eight of the 14 studies used only assessments that evaluated impairment-level constructs such as manual dexterity (Keyboard Personal Computer Style instrument, Baker & Redfern, 2005, used in Baker and Redfern, 2009, and the Box-and-Block Test, Cromwell, 1976, used in Canny et al., 2009); cognition (Loewenstein Occupational Therapy Cognitive Assessment; Itzkovich, Elazar, Averbuch, & Katz, 2000) and general intelligence (Pictorial IQ test; Hsu & Lu, 1995), both in Jang, Chern, and Lin (2009); grip strength (dynamometer; Lindstrom-Hazel, Kratt, & Bix, 2009); kinematics through motion monitoring of upper-extremity movement (May-Lisowski & King, 2008); and visual function (Brain Injury Visual Assessment Battery for Adults; Warren, 1998) used in Warren et al. (2008) . Although understanding impairment is indeed relevant to occupational performance, this level of assessment predominates the occupational therapy studies reviewed here. More important, information from this level of performance evaluation is not being translated into greater understanding of activity- and participation-level interventions. The authors of these studies have generally not made explicit how such assessments help understand the person–task–environment interaction that is at the heart of human occupation.

Only 2 of the 14 studies included environmental modifications. Darragh et al. (2008)  and Gentry (2008)  modified work and home environments, respectively, to improve their clients’ functional participation. To prevent work injuries, Darragh et al. (2008)  altered workstations to improve body position and ergonomic design of workstations for workers manufacturing microscopes. Gentry (2008)  modified clients’ home computers and gave them personal digital assistants to use at home to enhance their functional memory to perform everyday life tasks.

One of the most telling statistics about these articles is not reported in the tables. Five of the 14 studies did not acknowledge a funding source, 2 reported funding internal to their organization, 2 were student or PhD research, 1 was funded by a national society, 2 were funded by National Institutes of Health career development K-awards, 1 was funded by a government science award (Taiwan), and 1 was funded by a National Institute on Disability and Rehabilitation Research field-initiated award. Only the last three categories represent research proposals that receive significant external peer review. The career development awards are encouraging because these awards build the next cadre of researchers. However, building a systematic body of rehabilitation research requires investigation over multiple years, requiring sustained research funding not generally reflected in this group of articles.

Occupational therapy research in the areas of rehabilitation, disability, and participation and work and industry is showing progress toward meeting the Centennial Vision, with several studies focusing on context-based interventions that result in changes in client activity and participation. Activity- and participation-level outcome measures were the focus of several studies, yet most studies continue to use impairment-level measures as the primary outcome.

An area of concern for an intervention-focused profession such as occupational therapy is the paucity of rigorous effectiveness studies, that is, Level 1 research. In an era of increasing accountability and public reporting, clear evidence-based demonstrations of the effectiveness of occupational therapy research will be critical to the profession’s continued credibility and value. For example, Medicare is increasingly requiring public reporting of quality indicators (Rhoads, Konety, & Dudley, 2009). As we have noted previously, quality indicators require that we define and demonstrate the effectiveness of our interventions, deliver them consistently to all relevant clients, and routinely document their outcomes (Mallinson, Fischer, Rogers, Ehrlich-Jones, & Chang, 2009). These concerns are also reflected in the new priorities of the AOTA/American Occupational Therapy Foundation Research Advisory Panel (2010), which noted that in the next decade it will be “imperative that efficacy and effectiveness of occupational therapy interventions be ascertained” (p. 1).

We found evidence that a handful of rehabilitation studies focused on the effectiveness of occupational therapy in promoting occupational engagement and well-being. These studies highlight the importance of context in producing therapeutic change. Yet, a major concern remains that enough of these kinds of studies simply do not exist to build a coherent system of knowledge that serves as a foundation for advancing the Centennial Vision.

*indicates studies that were systematically reviewed for this article.