Manual physical assessment of spinal segmental motion: Intent and validity

Transcription

1 Available online at Manual Therapy 14 (2009) 36e44 Original article Manual physical assessment of spinal segmental motion: Intent and validity J. Haxby Abbott a, *, Timothy W. Flynn b, Julie M. Fritz c, Wayne A. Hing d, Duncan Reid e, Julie M. Whitman b a Centre for Physiotherapy Research, School of Physiotherapy, University of Otago, PO Box 56, Dunedin, New Zealand b Department of Physical Therapy, Regis University, Denver, CO, USA c Division of Physical Therapy, University of Utah, Salt Lake City, UT, USA d School of Physiotherapy, Auckland University of Technology, Auckland, New Zealand e Division of Rehabilitation and Occupation Studies, Auckland University of Technology, Auckland, New Zealand Received 19 March 2007; received in revised form 10 August 2007; accepted 13 September Abstract Validity of a clinical test can be defined as the extent to which the test actually assesses what it is intended to assess. In order to investigate the validity of manual physical assessment of the spine, it is therefore essential to establish what physical therapists intend to assess when they are applying these tests. The aims of this study were to (1) establish what manual physical therapists are intending to assess while applying passive intervertebral motion tests; and (2) examine the face validity and content validity for manual physical assessment of the spine. We surveyed 1502 members of the national manual physical therapist organisations of New Zealand and the United States of America using a web-based survey instrument. Sixty-six percent of 466 respondents believed passive accessory intervertebral motion (PAIVM) tests were valid for assessing quantity of segmental motion, and 76% believed passive physiologic intervertebral motion (PPIVM) tests were valid for assessing quantity of segmental motion. Ninety-eight percent of manual physical therapists base treatment decisions at least in part on the results of segmental motion tests. Quality of resistance to passive segmental motion was considered of greater importance than quantity of kinematic motion during PAIVM tests, while the quality of complex kinematic motion was considered of greater importance than quantity of displacement kinematics during PPIVM tests. Manual physical therapists accept the face validity of manual physical assessment of spinal segmental motion to a great extent, however a minority voice scepticism. Content validity is dominated by concepts of segmental kinematics and the forceedisplacement relationship. Intent of assessment does, however, vary widely between therapists. These data will inform the design of concurrent validity studies. Further work is recommended to increase consistency of intent, methodology and terminology in manual physical assessment of the spine. Ó 2007 Elsevier Ltd. All rights reserved. Keywords: Physical therapy techniques; Manual therapies 1. Introduction Expert clinicians believe that lumbar spinal segmental (intervertebral) motion is an important factor in low * Corresponding author. Tel.: þ ; fax: þ address: haxby.abbott@otago.ac.nz (J.H. Abbott). back pain (LBP) and that it can be assessed using physical examination procedures (Binkley et al., 1993). Some form of physical assessment of spinal motion is integral to almost all classification systems used by expert musculoskeletal physical therapists (PTs) to diagnose subgroups within non-specific spinal disorders (Paris, 1985; Nyberg, 1993; Delitto et al., 1995; Riddle, 1998; X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved. doi: /j.math

2 J.H. Abbott et al. / Manual Therapy 14 (2009) 36e44 37 Van Dillen et al., 1998; Laslett and van Wijmen, 1999; Petersen et al., 1999; Flynn et al., 2002; Cook et al., 2006). Two common manual physical examination procedures used by PTs to detect spinal segmental mobility are passive accessory intervertebral motion tests (PAIVMs) and passive physiological intervertebral motion tests (PPIVMs) (Abbott et al., 2005). The face validity of these manual physical examination tests for detecting spinal segmental mobility is not universally accepted. The content universe of properties of motion the assessor is expected to assess [i.e. all of what is entailed in the entity the test is intended to measure, and the relative importance of each component of that entity (Portney and Watkins, 1993)] is not known. The literature contains a very wide range of descriptors of what may be felt with manual segmental mobility assessment (Maher et al., 1998; Cook et al., 2006), although two studies suggest that the key concepts are quantity of motion and the forceedisplacement relationship (Binkley et al., 1993; Maher et al., 1998). Beyond this, little is known about the relative importance placed on the quantities of motion that may be assessed, or the various characteristics that may be perceived. The aim of this study was to investigate the face validity and content validity of manual assessment of spinal segmental motion. We aimed to examine: (1) the extent to which manual PTs accept the face validity of two common manual physical assessment procedures, PAIVMs and PPIVMs, for detecting spinal segmental motion; and (2) the relative importance of each component of the core content universe for these tests. 2. Methods The target population for this study was members of the New Zealand Manipulative Physiotherapists Association (NZMPA) and the American Academy of Orthopaedic Manual Physical Therapists (AAOMPT). We included all PTs who: (a) were financial members of these organisations in December 2005; and (b) had an active address on file with the organisation. This protocol was reviewed by the Institutional Review Board of Regis University, and was designated IRB Exempt. Development of the survey instrument was approached in three phases. The principal investigator (JHA) developed version 1 as part of earlier research (Abbott, 2005). Pilot data from that research informed version 2, which was further developed through consultation with the co-investigators and colleagues. We then designed version 3 using WebSurveyor Ò ( and pilot-tested the survey instrument on a small sample of manual PTs (n¼15). Based on feedback from the pilot test, we reworded questions and responses where necessary for the final version. We populated the WebSurveyor database with the list of addresses provided by NZMPA and AAOMPT. The WebSurveyor Ò program sent an invitation to eligible PTs. Respondents submitted completed surveys via WebSurveyor Ò, which logged the submission, added the responses to the results database, and withheld further reminder notices to the address of the respondent. Non-respondents were ed three reminder notices at 1-week intervals. Our examination of the extent of face validity was based on three questions. In the first question, participants were asked to choose (on a Likert scale) how accurate they believe each of the physical assessment procedures (PAIVMs and PPIVMs) are for estimating the quantity of movement present at a lumbar segment. In the second question, participants were asked if they believe that it is possible, from the clinical examination, to recognise restricted or excessive lumbar segmental motion. The third question relied on the assumption that participants demonstrated acceptance of face validity if they reported that, in clinical practice, they select different treatment options for patients with LBP on the basis of lumbar segmental motion findings. Table 1 Responses available to respondents for survey questions on intent of manual physical assessment and basis for comparative decisions Responses available for ranking: intent of manual physical assessment Responses available for ranking: basis for comparative decisions Position of the vertebrae, or normality of the position of vertebrae during motion Quantity of angle of spinal bending Quantity of translation of the vertebrae Quality of resistance (e.g. greater or lesser stiffness, forcedisplacement relationship) Quality of end-feel or tissue-feel (e.g. capsule, disc, muscle, ligament, bony, cartilage) Quality of the motion during the movement, or normality of the path of vertebral motion The patient s pain response (verbal or observed) The mobility you expect for that segmental level, taking into account the patient s age and/or body type The mobility you expect for that segmental level, compared to your experience of assessing the same segmental level in other patients and normal individuals The mobility you expect for that segmental level, compared to other segments above and below The patient s pain response (verbal or observed) to the procedure

3 38 J.H. Abbott et al. / Manual Therapy 14 (2009) 36e44 We then examined the two principal aspects of content validity: the content universe and the relative order of importance of each component of that content universe (Portney and Watkins, 1993). We identified, from the manual PT literature, key properties of motion within the content universe related to intent of manual segmental motion assessment and incorporated these as available responses to survey questions (Table 1). To establish the basis on which PTs make comparative decisions regarding these motion properties, we asked participants to rank the choices available from most important to least important. Respondents could leave blank any choice that did not represent their modus operandi. We captured other components of the content universe, and other bases for decision-making, by including an other response. Respondents choosing other were invited to submit a free-text response. We also included a request for further comments at the end of the survey instrument. These were analysed qualitatively by two investigators (WH, DR), and independently by another investigator (JHA) using Editing Analysis Style principles (Crabtree and Miller, 1992), until all major themes were identified. The minimum number of like responses necessary to constitute a theme was set at 5%. A hierarchy of themes was derived from the data: meta-themes, themes, and (where necessary) subthemes. The final themes were subsequently agreed by consensus between the three investigators. Descriptive statistics (number, percent) were used to summarise the data. For ranked responses, we counted the frequency each response was ranked at each ranking level, and present only the frequency they were ranked by respondents as most important (#1) and second most important (#2). We calculated the frequency and proportion of free-text responses in each theme which arose in the qualitative analysis. Table 2 Characteristics of respondents Respondents Total 466/1502 (31%) Country of PT practice USA 395 (86%) NZ 64 (14%) 2 (0.2%) Direct patient care >30 h/week 280 (61%) 21e30 h/week 65 (14%) 20 or less hours/week 115 (25%) Low back pain patients >30% of caseload 280 (61%) 10e30% of caseload 154 (34%) <10% of caseload 25 (5%) Entry level PT degree Certificate/diploma/other 66 (14%) Bachelors 212 (46%) Masters 156 (34%) Clinical doctorate (DPT) 27 (6%) Highest earned degree Certificate/diploma/other 71 (17%) Bachelors 100 (22%) Masters 170 (37%) Clinical doctorate 81 (18%) Research doctorate 38 (8%) Professional designations a ABPTS certified (e.g. OCS) 188 (48%) MNZMPA or FAAOMPT 165 (41%) Neither 152 (38%) PT, physical therapy or physiotherapy; USA, United States of America; NZ, New Zealand; DPT, Doctor of Physical Therapy degree; ABPTS, American Board of Physical Therapy Specialties; OCS, Orthopaedic Certified Specialist; NZMPA, fully qualified Member of the New Zealand Manipulative Physiotherapists Association; FAAOMPT, Fellow of the American Academy of Orthopaedic Manual Physical Therapists. a Percentages exceed 100% because respondents may be in more than one category. 3. Results Four hundred sixty-six (31%) of 1502 eligible PTs responded. Characteristics of the participants are described in Table 2. The majority of respondents (65.9%; 95% confidence interval 61.5%, 70.0%) believed that central postero-anterior PAIVMs are somewhat accurate or very accurate for estimating the quantity of movement present at a lumbar segment (for example identifying restricted, normal, excessive movement) (Table 3). A still higher proportion (76.2%; 72.1%, 79.8%) believe that PPIVMs are somewhat accurate or very accurate for estimating the quantity of movement present at a lumbar segment (Table 3). Most respondents (97.6%; 95.8%, 98.6%) report that they select different treatment options for patients with LBP at least partly on the basis of lumbar segmental motion findings. When asked to consider what they are intending to assess when performing central postero-anterior PAIVMs, the greatest proportion of respondents ranked the patient s pain response (verbal or observed) as most Table 3 Responses to the questions Indicate how accurate you believe [central P-A PAIVMs/PPIVMs] are for estimating the quantity of movement present at a lumbar segment (for example identifying restricted, normal, excessive movement) PAIVMs: number PPIVMs: number (%) a (%) b Not at all accurate 42 (9.2%) 20 (4.4%) Somewhat inaccurate 114 (24.9%) 89 (19.4%) Somewhat accurate 250 (54.7%) 279 (60.9%) Very accurate 51 (11.2%) 70 (15.3%) PAIVMs, passive accessory intervertebral motion tests; PPIVMs, passive physiological intervertebral. a Percent of valid data, 9 (1.9%) missing. b Percent of valid data, 8 (1.7%) missing.

4 J.H. Abbott et al. / Manual Therapy 14 (2009) 36e44 39 important (31.8% ranked this response #1; 18.8% ranked it #2). Quality of resistance (e.g. greater or lesser stiffness, forceedisplacement relationship) was the next most highly ranked (28.8 #1; 28.8 #2), followed by quantity of translation of the vertebrae (11.5; 13.7). Further results are presented in Fig. 1. When asked to consider what they intend to assess when they perform PPIVMs, quality of the motion during the movement (or normality of the path of vertebral motion) was ranked most important by respondents (29.7% ranked this response #1; 24.8% ranked it #2). Quantity of angle of spinal bending (e.g. flexion, extension, or sidebending) was the next most highly ranked (19.5 #1; 16.6 #2), followed by position of the vertebrae during motion (13.7; 9.8). Further results are presented in Fig. 2. How respondents judge spinal mobility during passive intervertebral motion assessment procedures, is presented in Fig. 3. The free-text responses offered by participants fell into two meta-themes: justification and clarification. The themes revealed for PAIVMs and PPIVMs were not dissimilar, so were collapsed into common themes. Results from the 73 free-text responses offered are listed in Table 4. Regarding what manual PTs intend to assess using manual tests of segmental motion, comments suggested that interaction between properties (or domains) are assessed by some respondents: A general sense of movement quality related to quality determinants such as palpable resistance, muscle spasm, and related to other patient psycho-physiological responses. Property Position of vertebra Quantity of angle Quantity of translation Quality of resistance Quality of end-feel Quality of motion path Pain response Ranked #1, # Fig. 1. Responses to the question Consider what you are intending to assess when you perform central P-A PAIVMs. Using the drop-down boxes, please rate the following choices in rank order, from most important to least important. Data represent the proportion (in percent) of manual physical therapists who ranked each response as of the greatest importance (dark red bar) and second greatest importance (light blue bar) during assessment of spinal segmental motion. Four hundred and fifty-nine (98.5% of) respondents gave at least 1 ranked response, 452 (97% of) respondents gave 2 or more ranked responses. Data were missing for 7 (1.5%) participants. Property Position of vertebra Quantity of angle Quantity of translation Quality of resistance Quality of end-feel Quality of motion path Pain response Ranked #1, # The pain-range behavior which I ll define as the where in range is the pt s concordant sign reproduced and with how much force. Several respondents were sceptical about the validity of manual physical assessment of spinal segmental motion: I do not use this technique in general. There is very weak EBP Very rarely use them. The literature indicates they are of little value with poor reliability to determine a level let alone anything else of use Fig. 2. Responses to the question Consider what you are intending to assess when you perform central PPIVMs. Using the drop-down boxes, please rate the following choices in rank order, from most important to least important. Data represent the proportion (in percent) of manual physical therapists who ranked each response as of the greatest importance (dark red bar) and second greatest importance (light blue bar) during assessment of spinal segmental motion. Four hundred and fifty-one (96.8% of) respondents gave at least 1 ranked response, 440 (94.4% of) respondents gave 2 or more ranked responses. Data were missing for 15 (3.2%) participants. Judgement made by mobility expected compared to: That segmental level considering age & body type That segment allevel cf. other patients & normals Segments above & below within the same patient Pain response Ranked #1, # Fig. 3. Responses to the question Consider how you judge spinal mobility. Do you base your judgments on any of the following? Using the drop-down boxes, please rate the following choices in rank order, from most important to least important. Data represent the proportion (in percent) of manual physical therapists who ranked each response as of the greatest importance (dark red bar) and second greatest importance (light blue bar) during assessment of spinal segmental motion. Four hundred and fifty-nine (98.5% of) respondents gave at least 2 ranked responses. Data were missing for seven (1.5%) participants.

5 40 J.H. Abbott et al. / Manual Therapy 14 (2009) 36e44 Table 4 Themes identified from other responses to the questions Consider what you are intending to assess when you preform. PAIVMs and PPIVMs Theme a Number (%) b Examples from the data Interaction of domains 12 (16%) Pain response combined with quantity of motion Relationship of pain and resistance Here I am looking for different things if the patient s condition is acute or more chronic Muscle-related 7 (10%) Influence of muscle length on vertebral movement Paraspinal muscle reactivity or reaction Symptom-related 5 (7%) Ability to reproduce patient s symptoms Scepticism 4 (5%) I don t use this technique in general. There is very weak EBP The literature indicates they are of little value with poor reliability. PAIVMs, passive accessory intervertebral motion tests; PPIVMs, passive physiological intervertebral motion tests. a Two other common themes (movement-related and position by palpation) are not listed as they replicate responses available in the survey instrument. b Number (%) of 73 comments represented by theme. Percentages do not sum to 100, because some comments either could not be classified by theme, or represented responses available in the survey questions (Table 1). Regarding how manual PTs make comparative judgments about the motion of a spinal segment, one main theme was identified from the 37 free-text responses offered by respondents (Table 5). Respondents emphasised the importance of the patient s history and other clinical findings, indicating that the influence of pretest probability is taken into account: Expected mobility for that segment based on the patient s history of injury and course of symptom progression/regression. Patients response to active movement, observation skills patterns of movement and postural influences, i.e. descrepencies [sic] in anatomy, weak muscles etc. Comments offered in the final free-text question revealed three main themes (Table 6). Among the 144 free-text responses, many respondents commented that manual tests of segmental motion are but a part of a more comprehensive set of data collected to inform clinical decision-making: PPIVM and PAIVM must be used along with other tests to rule in or out diagnostic hypotheses that were formed from the history and subjective report. One test does not stand alone. The theme of scepticism regarding the reliability and validity of manual tests of segmental motion continued:.i believe the literature to date does not support this approach. although some acknowledged evidence supporting their valid use: I rarely use PPIVMs now. Unreliable and invalid. I use them only to identify the anatomical [sic] segment once I have chosen to perform a manipulative procedure. I use PAIVMs primarily for pain response (provocation or abolition). I now consider (on the basis of current evidence) that PAIVMs may have value in a composite of clinical findings for identifying patients most likely to respond to manipulation (Flynn, Childs, Fritz papers). As noted in the lit [literature] the reliability of PPIVM s and PAIVM s is poor however combined with other subjective and objective data we can subgroup LBP patients and apply CPR s [clinical prediction rules]. 4. Discussion Our data indicate that most (66e76%) manual PT respondents have faith in the accuracy of manual segmental mobility assessment procedures for detecting quantity of spinal segmental motion. Few (5e9%) expressed scepticism, or rated the procedures as not at all accurate. The overwhelming majority (98%) base patient treatment choices at least partly on the basis of lumbar segmental motion assessment findings. These Table 5 Themes identified from other responses to the question Consider how you judge spinal mobility. Do you base your judgments on. Theme Number (%) a Examples from the data Taking other history and physical findings into account (sum of subthemes) b 22 (59%) Based on medical/surgical history and what patient presents in his subjective history/complaints Patient history/description of activities/positions that produce the pain/catching; irritability of presentation Range of motion testing; observing quality and quantity of motion a Number (%) of 37 comments represented by theme. Percentages do not sum to 100, because some comments could not be classified by theme, or represented responses available in the survey questions (Table 1). b Subthemes identified were: History examination; Symptoms (including reproduction of); physical examination items.

6 J.H. Abbott et al. / Manual Therapy 14 (2009) 36e44 41 Table 6 Themes identified from responses to the question Would you like to comment on any aspect of the use of PAIVM or PPIVM assessments for the examination of the spine? Theme Number (%) a Examples from the data Part of a multi-factorial process 49 (34%) I view PPIVM and PAIVM as part of a comprehensive examination process in orthopedic [sic] physical therapy. Findings are taken within the context of all physical examination findings, as well as the subjective portion of the patient examination I do it differently 14 (10%) I perform PAIVMs superior and inferior directions to access facet restrictions, not in a PeA direction Generally don t use them due to McKenzie training Scepticism 13 (9%) Questionable interrater reliability I rarely use PPIVMs now. Unreliable and invalid Experience is necessary to use these tests 10 (7%) I also think that the ability of a therapist to consistently use these tools increases there [sic] reliability and validity Essential to the clinical examination 7 (5%) A spinal evaluation is not complete unless PPIVM and/or PAIVM are performed! PAIVM, passive accessory intervertebral motion; PPIVM, passive physiological intervertebral motion. a Number (%) of 144 comments represented by theme. Percentages do not sum to 100, because some comments could not be classified by theme, or represented responses available in the survey questions (Table 1). findings are consistent with previous research indicating that segmental motion is an important clinical finding in the examination and diagnosis of patients with spinal pain (Binkley et al., 1993; Cook et al., 2005a, 2006), and PTs are confident in diagnosing segmental motion disorders (Cook et al., 2005b). Our data concur with Cook and colleagues (2005b), who found that over 70% of manual PTs in their Dephi study often or very often use palpable segmental mobility abnormalities in the diagnosis of spinal instability (Cook et al., 2005b). We therefore conclude that acceptance of the face validity of manual segmental mobility assessment procedures is strong among manual PTs. Our data indicate that, overall, quantity of motion and quality of the forceedisplacement relationship were the dominant concepts within the content universe of spinal segmental motion, concurring with earlier research (Binkley et al., 1993; Maher et al., 1998; Cook et al., 2006). Few respondents ranked other properties of segmental motion as highly important, for either PAIVM or PPIVM testing, which supports the view that the content universe of what is intended to be assessed during these manual tests was adequately represented by the responses available (see Table 1). Within these responses, however, considerable variability was seen in the data. Patients pain response to PAIVM testing was rated of highest importance by more respondents (32%) than rated any of the motion properties highest, which is consistent with research indicating that these assessment procedures are accurate for the detection of a painful segment (Phillips and Twomey, 1996). A large proportion of respondents (58%: equally split between #1 and #2 rankings) intend to assess the quality of resistance to movement (i.e. greater or lesser stiffness; the force-displacement relationship) during PAIVM testing, as found by (Maher et al., 1998), and supported by experimental research using a proxy criterion standard for stiffness (Chiradejnant et al., 2003a). A wide variety of other motion properties were ranked highly by respondents. Manual PTs also commonly use PAIVM tests to assess quantity of translation and the quality of the end-feel of segmental motion. Few studies have measured sagittal translation motion, however, those that have report that the magnitude of motion is small (Lee and Evans, 2000), may not be manually perceptible (Lee and Evans, 1997), and may be in the opposite direction to that which some manual PT theories might predict. Posterior translation seems to occur because of the spine behaving like a flexible beam undergoing three-point bending, rather than anterior translation of one vertebra on the neighbouring vertebra below (Lee and Evans, 1997). Although these in vitro and in vivo laboratory studies findings dispute the face validity of manual segmental mobility assessment, two recent criterion-related validity studies independently concluded that PAIVMs may possess concurrent validity for detecting excessive sagittal translation (Abbott et al., 2005; Fritz et al., 2005a). Despite differing methodology and populations, their results were highly consistent, which provides rigorous independent validation (Abbott, 2007). Although in one of the studies (Fritz et al., 2005a) the likelihood ratio for a positive test (LRþ) was not statistically significant, it appears likely that this may be due to type 2 error as the other study (Abbott et al., 2005) with a larger sample size found almost exactly the same LRþ estimate, and was statistically significant. Similarly, in one of the studies (Abbott et al., 2005) the likelihood ratio for a negative test (LR-) was not statistically significant, however, again this is likely to be type 2 error, as the study with the higher prevalence of excessive sagittal translation (Fritz et al., 2005a) found a highly comparable LRestimate and was statistically significant. The magnitude

7 42 J.H. Abbott et al. / Manual Therapy 14 (2009) 36e44 of these associations is, however, small, indicating that: (a) a single clinical examination procedure in isolation has only weak diagnostic value; and possibly also that: (b) the kinematic properties measured as a criterion in these studies may not adequately represent the content universe of what manual PTs are assessing while performing these examination procedures. Research indicates that sagittal rotation occurs during PAIVM testing (Powers et al., 2003; Kulig et al., 2004, 2007; Beneck et al., 2005; Lee et al., 2005), although they are noted to be small, leading some authors to speculate that accurate detection by manual means would unlikely (Lee and Evans, 1997). Lee and Evans (1997) also note that the vertebrae rotate and translate simultaneously, and suggest it unlikely the assessing therapist would be able to discriminate the component movement properties. Limited motion will certainly be of a small magnitude, and it is unclear whether excessive sagittal rotation is an important factor associated with LBP (Abbott et al., 2005, 2006). Our data, however, suggest that very few manual PTs intend to assess sagittal rotation during PAIVM testing, which diminishes the external validity of such speculation. There have been few technical studies of PPIVMs, however, excessive sagittal translation motion has been shown to be more readily detected by flexione extension in the side-lying position than by standing flexioneextension (Wood et al., 1994), and there is preliminary evidence that PPIVMs performed in side-lying may have moderate validity for detecting excessive sagittal translation, however, further research must be done to confirm these findings (Abbott and Mercer, 2003; Abbott et al., 2006). No studies have yet examined complex kinematics during spinal segmental mobility testing. The quality of the path of motion perceived by the assessor during PPIVM testing was ranked of highest importance by respondents, both in terms of #1 ratings and the aggregate of #1 and #2 ratings. This indicates that complex kinematics were considered more important than simple displacement kinematics (such as quantity of angledsecond most highly rateddand quantity of translation). This proposition will pose technical difficulties to researchers choosing a reference standard for assessing the criterion-related validity of PPIVM testing, however, emerging evidence both supports the relative importance of complex kinematics over simple displacement kinematics in patients with LBP, and offers technical solutions to measuring it (Abbott, 2005; McCane et al., 2006; Teyhen et al., 2007). The data we have provided on manual PTs intent will inform researchers in the design and interpretation of criterion-related concurrent validity studies, helping to ensure a match between intent and an appropriate reference standard. The present data indicate that manual PTs intent of assessment encompasses more than one domain, therefore when interpreting the results of criterion-related concurrent validity studies readers should take this fact into account. Criterion-related validity studies may underestimate validity if they examine only one domain, such as sagittal translation, without examining other important domains, such as stiffness. Estimates of the actual validity must take into account criterion-related validity of the important biomechanical domains therapists intend to assess concurrently (such as displacement, aberrant path of motion, and stiffness) as well as predictive validity. With little evidence of the validity of manual physical assessment procedures, our data suggest that some manual PTs may have turned to research investigating their reliability as a proxy measure to support or refute face validity. We suggest that it is a poor choice of proxy for three principal reasons. Firstly, we refer the reader to (Wainner, 2003) for enlightening data and argument on why low reliability may not preclude diagnostic or prognostic utility, although it must be noted that the data Wainner presents are single estimates from different studies, indicating that further research is required to adequately support the theory he presents. Secondly, the threshold for how reliable a test must be in order to be useful depends on several factors, not least being the consequences of an erroneous result (Rothstein, 2001; Wainner, 2003). And thirdly, usefulness (i.e. validity) is a more important property of a test than reliability, but is generally more difficult to study (Rothstein, 2001). There is growing evidence that manual physical assessment procedures are useful, in that they demonstrate predictive validity both for the purpose of making treatment decisions and for outcome, particularly when combined with other factors in a multivariate decision rule (Chiradejnant et al., 2002; Flynn et al., 2002; Fritz et al., 2003, 2004, 2005b; Childs et al., 2004; Hicks et al., 2005), although not all studies support these propositions (Haas et al., 2003; Chiradejnant et al., 2003b). Our data, indicating that the majority of manual PTs believe manual physical assessment procedures are moderately (somewhat) accurate, and overwhelmingly use the results of those procedures to make treatment decisions, indicates that manual PTs practice is consistent with the available evidence on usefulness. Our data also indicate that manual PTs recognise that the validity of these tests is not strong enough to stand alone, so use them in conjunction with other observations, tests and measures. A plausible explanation for poor reliability, according to Maher et al. (1998), may be that therapists knowledge of what biomechanical properties they are able to assess, and what properties are important in LBP, is inadequate. Terminology is inconsistent and often jargon-filled (Maher et al., 1998). We concur with this view, and believe that this inadequate knowledge base and terminology may explain the considerable

8 J.H. Abbott et al. / Manual Therapy 14 (2009) 36e44 43 variation in the data from the respondents to this study. Consistent methodology and terminology grounded in sound biomechanical and clinical evidence may facilitate communication between teacher and student; therapist and patient; clinician and researcher; and between members of the interdisciplinary health-care team. Conclusions drawn from these results must take into account the strengths and limitations of the data. The data represent a large sample of clinically active manual physical therapists across two countries, suggesting international generalisability of the results, however, the small sample from New Zealand (due to its smaller population) does not permit subgroup comparisons. Respondent bias may have inflated our estimates of face validity, because it is likely the 31% of eligible manual PTs who responded have a greater interest in, and therefore belief in, the validity of the manual physical assessment techniques in question. As we were not able to capture a sample of non-respondents, we cannot discount the possibility that the responses of participants may differ from those of non-participating manual PTs. However, the possibility of respondent bias is mitigated by evidence of relatively high levels of scepticism regarding these assessment techniques. The use of prepared responses that respondents were asked to rank in order of importance may have led to bias, however, this is mitigated by the fact that respondents did not offer a large number of free-text responses, and very few of these were ranked highly by respondents. 5. Conclusions Manual physical therapists attribute moderate to strong face validity to manual passive assessment techniques, both for assessing spinal segmental motion and as a basis for determining what type of intervention is indicated. There remains some scepticism regarding the validity of manual passive assessment techniques, despite growing evidence of their concurrent and predictive validity. This scepticism is associated with the perception of low reliability. Our data concerning content validity indicate that manual passive assessment techniques are intended to provide the therapist with information on the location of a painful segment, the quality of the forceedisplacement relationship, the quantity of segmental motion, and to a much lesser extent the underlying patho-anatomical basis for any abnormality detected. These results will inform the design and interpretation of criterion-related validity studies. We recommend teachers and practitioners of manual physical assessment techniques adopt scientifically sound and consistent methodology and terminology in describing the intent and content of these assessment procedures. Acknowledgments Thanks AAOMPT, NZMPA, in particular Ken Olsen and Vicki Reid; Dr Darren Rivett; and all the manual physical therapists who took the time to complete and submit this survey. References Abbott JH. Accuracy in the diagnosis of lumbar segmental mobility disorders. Ph.D. Thesis, University of Otago, Dunedin, Abbott JH. Passive intervertbral motion tests for diagnosis of lumbar instability. Australian Journal of Physiotherapy 2007;53(1):66. Abbott JH, Mercer SR. Lumbar segmental hypomobility: criterionrelated validity of clinical examination items (a pilot study). New Zealand Journal of Physiotherapy 2003;31(1):3e9. Abbott JH, McCane B, Herbison P, Moginie G, Chapple C, Hogarty T. Lumbar segmental instability: a criterion-related validity study of manual therapy assessment. BMC Musculoskeletal Disorders 2005;6:56. Abbott JH, Fritz JM, McCane B, Shultz B, Herbison P, Lyons B, et al. Lumbar segmental mobility disorders: comparison of two methods of defining abnormal displacement kinematics in a cohort of patients with non-specific mechanical low back pain. BMC Musculoskeletal Disorders 2006;7:45. Beneck GJ, Kulig K, Landel RF, Powers CM. The relationship between lumbar segmental motion and pain response produced by a posterior-to-anterior force in persons with nonspecific low back pain. Journal of Orthopaedic & Sports Physical Therapy 2005;35(4):203e9. Binkley J, Finch E, Hall J, Black T, Gowland C. Diagnostic classification of patients with low back pain: report on a survey of physical therapy experts. Physical Therapy 1993;73(3):138e50. Childs JD, Fritz JM, Flynn TW, Irrgang JJ, Johnson KK, Majkowski GR, et al. A clinical prediction rule to identify patients with low back pain most likely to benefit from spinal manipulation: a validation study. Annals of Internal Medicine 2004;141(12): 920e8. Chiradejnant A, Latimer J, Maher CG, Stepkovitch N. Does the choice of spinal level treated during posteroanterior (pa) mobilisation affect treatment outcome? Physiotherapy Theory and Practice 2002;18(4):165e74. Chiradejnant A, Maher CG, Latimer J. Objective manual assessment of lumbar posteroanterior stiffness is now possible. Journal of Manipulative and Physiological Therapeutics 2003;26(1):34e9. Chiradejnant A, Maher CG, Latimer J, Stepkovitch N. Efficacy of therapist-selected versus randomly selected mobilisation techniques for the treatment of low back pain: a randomised controlled trial. Australian Journal of Physiotherapy 2003;49(4):233e41. Cook C, Brismee JM, Fleming R, Sizer Jr PS. Identifiers suggestive of clinical cervical spine instability: a delphi study of physical therapists. Physical Therapy 2005;85(9):895e906. Cook C, Brismee JM, Sizer PS. Factors associated with physiotherapists confidence during assessment of clinical cervical and lumbar spine instability. Physiotherapy Research International 2005;10(2): 59e71. Cook C, Brismee JM, Sizer Jr PS. Subjective and objective descriptors of clinical lumbar spine instability: a delphi study. Manual Therapy 2006;11(1):11e21. Crabtree BF, Miller WL. Doing qualitative research. Newbury Park, London: Sage; p. 18e20. Delitto A, Erhard RE, Bowling RW. A treatment-based classification approach to low back syndrome: identifying and staging patients for conservative treatment. Physical Therapy 1995;75(6):470e85.

9 44 J.H. Abbott et al. / Manual Therapy 14 (2009) 36e44 Flynn T, Fritz J, Whitman J, Wainner R, Magel J, Rendeiro D, et al. A clinical prediction rule for classifying patients with low back pain who demonstrate short-term improvement with spinal manipulation. Spine 2002;27(24):2835e43. Fritz JM, Whitman JM, Flynn TW, Wainner RS, Childs JD. Clinical factors related to the failure of individuals with low back pain to improve with spinal manipulation. Journal of Orthopaedic & Sports Physical Therapy 2003;33:A4e5. Fritz JM, Whitman JM, Flynn TW, Wainner RS, Childs JD. Factors related to the inability of individuals with low back pain to improve with a spinal manipulation. Physical Therapy 2004;84(2):173e90. Fritz JM, Piva SR, Childs JD. Accuracy of the clinical examination to predict radiographic instability of the lumbar spine. European Spine Journal 2005;14(8):743e50. Fritz JM, Whitman JM, Childs JD. Lumbar spine segmental mobility assessment: an examination of validity for determining intervention strategies in patients with low back pain. Archives of Physical Medicine and Rehabilitation 2005;86(9):1745e52. Haas M, Groupp E, Panzer D, Partna L, Lumsden S, Aickin M. Efficacy of cervical endplay assessment as an indicator for spinal manipulation. Spine 2003;28(11):1091e6 [discussion 6]. Hicks GE, Fritz JM, Delitto A, McGill SM. Preliminary development of a clinical prediction rule for determining which patients with low back pain will respond to a stabilization exercise program. Archives of Physical Medicine and Rehabilitation 2005;86(9):1753e62. Kulig K, Landel R, Powers CM. Assessment of lumbar spine kinematics using dynamic mri: a proposed mechanism of sagittal plane motion induced by manual posterior-to-anterior mobilization. Journal of Orthopaedic & Sports Physical Therapy 2004; 34(2):57e64. Kulig K, Powers CM, Landel RF, Chen H, Fredericson M, Guillet M, et al. Segmental lumbar mobility in individuals with low back pain: in vivo assessment during manual and self-imposed motion using dynamic mri. BMC Musculoskeletal Disorders 2007;8:8. Laslett M, van Wijmen P. Low back and referred pain: diagnosis and a new system of classification. New Zealand Journal of Physiotherapy 1999;27(2):5e14. Lee R, Evans J. An in vivo study of the intervertebral movements produced by posteroanterior mobilization. Clinical Biomechanics (Bristol, Avon) 1997;12(6):400e8. Lee RY, Evans JH. The role of spinal tissues in resisting posteroanterior forces applied to the lumbar spine. Journal of Manipulative & Physiological Therapeutics 2000;23(8):551e6. Lee RY, McGregor AH, Bull AM, Wragg P. Dynamic response of the cervical spine to posteroanterior mobilisation. Clinical Biomechanics (Bristol, Avon) 2005;20(2):228e31. Maher CG, Simmonds M, Adams R. Therapists conceptualization and characterization of the clinical concept of spinal stiffness. Physical Therapy 1998;78(3):289e300. McCane B, King TI, Abbott JH. Calculating the 2d motion of lumbar vertebrae using splines. Journal of Biomechanics 2006;39:2703e8. Nyberg R. Clinical assessment of the low back: active movement and palpation testing. In: Basmajian JV, Nyberg R, editors. Rational manual therapies. Baltimore: Williams & Wilkins; p. 97e140. Paris SV. Physical signs of instability. Spine 1985;10(3):277e9. Petersen T, Thorsen H, Manniche C, Ekdahl C. Classification of nonspecific low back pain: a review of the literature on classifications systems relevant to physiotherapy. Physical Therapy Reviews 1999;4(4):265e81. Phillips DR, Twomey LT. A comparison of manual diagnosis with a diagnosis established by a uni-level lumbar spinal block procedure. Manual Therapy 1996;1(2):82e7. Portney LG, Watkins MP. Foundations of clinical research: applications to practice. East Norwark, CT: Appleton & Lange; Powers CM, Kulig K, Harrison J, Bergman G. Segmental mobility of the lumbar spine during a posterior to anterior mobilization: assessment using dynamic mri. Clinical Biomechanics 2003;18(1): 80e3. Riddle DL. Classification and low back pain: a review of the literature and critical analysis of selected systems. Physical Therapy 1998;78(7):708e37. Rothstein JM. Sick and tired of reliability? Physical Therapy 2001;81(2):774e5. Teyhen DS, Flynn TW, Childs JD, Abraham LD. Arthrokinematics in a sub-group of patients likely to benefit from a lumbar stabilization exercise program. Physical Therapy 2007;87(3):313e25. Van Dillen LR, Sahrmann SA, Norton BJ, Caldwell CA, Fleming DA, McDonnell MK, et al. Reliability of physical examination items used for classification of patients with low back pain. Physical Therapy 1998;78(9):979e88. Wainner RS. Reliability of the clinical examination: how close is close enough? Journal of Orthopaedic & Sports Physical Therapy 2003;33(9):488e91. Wood KB, Popp CA, Transfeldt EE, Geissele AE. Radiographic evaluation of instability in spondylolisthesis. Spine 1994;19(15): 1697e703.