Outcomes Measurement for Patients With Low Back Pain Linda Resnik Ed Dobrykowski Outcomes tracking provides a systematic method of monitor- ing treatment effectiveness and efficiency. A familiarity with outcome measures for the patient with low back pain is very important for clinicians working in orthopaedic settings, where patients with lumbar pain are prevalent. The clinician must be able to evaluate and choose appropriate measurement tools, and understand the clinical meaning of measurements to suc- cessfully employ these instruments. The purposes of this article are to review measurement instruments and to offer practical guidelines for selection and use of outcome measures for this population. The reliability, validity, sensitivity to change, and utility of common outcome measures are discussed. An overview of generic, disease-specific, and patient-specific tools is provided, vwith specific commentary on the use of the SF-36, SF-12, Oswestry, Roland Morris, and patient-specific tools. Practical guidelines for utilizing outcome measures in clinical practice and the overall benefits of outcomes tracking are highlighted. Tin here is a growing interest among healthcare providers using outcome measures to direct quality improve- ment initiatives (Doyle, 1995 Sakallaris, Jastremski, & Von Rucdon, 2000), assist in development of clinical guidelines (Crosson, 1995), and make decisions about in- dividual palients {Basmajian, 1995 Fiitz & Irrgang, 2001 Stratford & Binklcy, 1995 Zander, 1998). This trend is consistent with the final step of an evidence- based approach to practice (EBP) which is evaluation of one's clinical petformance (Deaton. 2001 DeLise & Leasure, 2001 Lewis & Latney, 2002 Sackett, Rosenberg, Gray. Haynes. & Richardson, 1996). Although many clinicians recognize the importance and value of outcomes measurement, they lack the nec- essary tiaining and experience to successfully incorpo- rate these measures into their practice (Basmajian, 1995). Clinicians should be able evaluate and choose appropriate outcome measures for their population of patient-s, make inferences about changes in measure- ment scores that occur during treatment, and use this information for clinical decision making and quality improvement. However, prior studies show that providers have concerns about selection, administration, scoring, interpretation, and documentation of outcomes that must be addressed to facilitate their usage (Basmajian, 1995). A familiarity with outcome measures for the patient with low back pain (LBP) is essential for clinicians in working in orthopaedic settings where patients with lum- bai- spine pain often comprise a substantial percentage of the caseload. The purposes of this article are lo discuss selection of outcome measures for patients with low hack pain, and to provide guidance for clinical inteipretation of outcomes information in this population. LL Outcomes in Patients With Low Back Pain Defining outcomes that are meaningful to stakeholders is a ciitical first step in outcomes measurement, A central challenge exists in that the definition of treatment suc- cess depends upon the stakeholder's perspective. Each stakeholder maintains his or her view on what constitutes a "good" or a meaningful result of care. From the patient and physician's perspective, the "good" outcome is often considered symptom relief (Grimmer, Sheppard, Pitt, Magarey, & Trott, 1999 Melles, Mclntosh, & Hall, 1995). Employers believe that a positive outcome is defined by theii" injured employee's return to work. Payors of health- care have still another viewpoint and regard successful outcomes as cost-efficient patient management and pa- tient satisfaction (Grimmer et al., 1999). Traditionally, clinicians have utilized changes in physi- cal impairments, such as spinal range of motion, abdom- inal strength, and straight leg raise as measurements of treatment progress and success. Studies of physical ther- apy pioviders show that they define "good outcomes" as the learning of long-term management strategies, relief of Linda Resnik. PhD, PT, OCS, Brown University. Providence, Rl. Ed DobTYkowski, PT, MHS, ATC/L, Multicare Health Systems, Tacoma, WA. 1 4 Orthopaedic Nursing ^ January/February 2005 *" Volume 24 ^ Number 1

symptoms, and improved function (Grimmer et al., 1999). Although these quality-of-life outcomes appear to be val- ued by providers of orthopaedic care, many clinicians do not know how to quantify or measure them (Basmajian, 1995). With a variety of outcomes and measures to choose from, which should clinicians utilize in patients with low back disorders, and how should they be used? We will now examine these outcome measures in further detail. Return to Work An employee's return to work after an episode of low back pain is an outcome that is highly valued by patients, em- ployers, payors, and society. Clinicians ii-equently include diminished work status among the patient problem list and include return to work as one of the treatment goals. However, the measurement of the patient's return to work as an outcome of healthcare has numerous limitations, which suggest that return to work may not be a valid in- dicator of improvement in patients with low back pain (Feuerstein & Beattie, 1995 Hunter, Shaha, Flint, & Tracy. 1998). Because improvement of function and quality of life is not necessarily related to work status, use of return to work as an indicator of treatment success is problematic. Occasionally, patients have made improvement but are not working for reasons unrelated to their low back pain. Return to work status fails to account for economic is- sues of job availability, homemaker and student status, retirement age, or those patients who remain on the job despite impairment (Beattie & Maher, 1997). On the opposite end of the spectrum, some patients never leave work, and others resume working while still experiencing limitations in job roles, productivity, and task performance. In these instances, gross measurement of return to work status does not measure on the job im- pact of low back pain (Lerner et al., 2001). Return to work has been shown to be affected by so- cioeconomic characteristics, economic incentives, job characteristics, and employment status (Baldwin, Johnson, & Butler, 1996 Dworkin, Handlin, Richlin, Brand, & Vannucci, 1985 Teasell & Bombardier, 2001). Patients whose jobs require a high physical workload are less likely to return to work following back injury (Fransen et al., 2002 Krause, Dasinger, Deegan, Rudolph, & Biand, 2001 Leavitt, 1992 Teasell & Bombardier, 2001). Return to work has been associated with sex, age, and time off from work (Beissner, Saunders, & McManis, 1996 Mayer, Gatchel, & Evans, 2001), as well as time before beginning treatment (Infante-Rivard & Lortie, 1996). Social factors including marital status and number of children have also been shown to be associated with the rate of return to work (Haldorsen, Indahl, & Ursin, 1998 Lehmann, Spratt, & Lehmann, 1993). Several authors have repotied a relationship between return to work and workers' compensation status (Atlas et al, 2000 Baldwin et al., 1996 Beissner et al., 1996 ToUison, 1993). Physical Impairments Impairments are defined as abnoimalities of stmcture or function, indicated by signs and symptoms (American Physical Therapy, 200i). Historically, clinicians have relied on measures of im- pairment (e.g., spinal range of motion, muscle strength), believing that treatment directed toward minimizing im- pairments will reduce the patient's pain and disability (Beattie & Maher, 1997). There are several limitations of using impairments as outcome measures, including lack of standardization of impairments (Fitzgerald, McClure, Beatiie, & Riddle, 1994 Lindsay, Meeuwisse, Mooney, & Summersides, 1995: Strender, Sjoblom, Sundell, Ludwng. & Taubo, 1997) and poor reliability of commonly used impairment meas- urements (Egan, Cole. & Twomey, 1996 Laslett & Williams, 1994 Levangie. 1999 Meijne. van Neetbos. Aufdemkampe, & van der Wurff. 1999 Potter &i Rothstein. 1985 Van Dillen et al., 1998). Physical impairment and physical functioning appear to be separate constructs that do not necessarily have a straightforward relationship (Jette, 1995). Physical impairment and physical functioning appear to be separate constructs that do not necessarily have a straightforward relationship (Jette, 1995). Furthermore, some measurements of impairment are not associated with patient fimction or disability, bringing into question their meaningfulness as outcome measures. For example. range of motion, and straight leg raise have been shown to have a poor correlation with disability (Hazard, Haugh, Green, & Jones, 1994 Nattrass, Nitschke, Disler, Chou.&Ooi, 1999). Further research is needed to determine which meas- urements of impairment are reliable and valid and may predict the risk for functional limitation and disability. As reliable and valid physical impairment measures are identified, the clinician should consider their use when treating patients with low back pain syndromes. Health-Related Quality of Life Health-related quality of life (HRQL) measures multiple domains of health that often inciude physical, psycholog- ical, emotional, and social dimensions (Jftte, 1993). HRQL instruments are tj^sically questionnaires, com- pleted by patients, which report their experience of what activities they can do, how often they can do them, and the level of difficulty they have peribiming them. HROL instruments have been widely recommended as an out- come measutement for patients with low back pain (Delitto. 1994 Deyo et al,, 1998 Enebo. 1998 Jette, 1995). In the literature and within the clinic, the term HRQL is often used interchangeably witb the terms func- tional status, health status, and health outcomes (Coons, Rao, Keininger, & Hays, 2000 Jette, 1993). A common misperception is that patients' responses to HRQL questionnaires are subjective and unreliable, be- cause patients are reporting their own personal subjective experiences. Clinicians have a tendency to view their "ob- jective" clinical measurements of impaiiment as more Orthopaedic Nursing ' January/February 2005 * Volume 24 ^ Number 1 1 5