Physiother Res Int. 2018 (Mar 14) [Epub] ~ FULL TEXT
Mark W. Werneke, Susan Edmond, Michelle Young, David Grigsby, Brian McClenahan, Troy McGill
Doctoral Programs in Physical Therapy,
Rutgers, The State University of New Jersey,
Newark, NJ, USA
BACKGROUND AND PURPOSE: Physiotherapy has an important role in managing patients with non-specific low back pain who experience elevated psychosocial distress or risk for chronic disability. In terms of evidence-based physiotherapy practice, cognitive-behavioural approaches for patients at high psychosocial risk are the recommended management to improve patient treatment outcomes. Evidence also suggests that directional preference (DP) is an important treatment effect modifier for prescribing specific exercises for patients to improve outcomes. Little is known about the influence of treatment techniques based on DP on outcomes for patients classified as high psychosocial risk using the Subgroups for Targeted Treatment (STarT Back Screening Tool). This study aimed to examine the association between functional status (FS) at rehabilitation discharge for patients experiencing low back pain classified at high STarT psychosocial risk and whose symptoms showed a DP versus No-DP.
METHODS: High STarT risk patients (n = 138) completed intake surveys, that is, the lumbar FS of Focus On Therapeutic Outcomes, Inc., and STarT, and were evaluated for DP by physiotherapists credentialed in McKenzie methods. The FS measure of Focus On Therapeutic Outcomes, Inc., was repeated at discharge. DP and No-DP prevalence rates were calculated. Associations between first-visit DP and No-DP and change in FS were assessed using univariate and multivariate regression models controlling for 11 risk-adjusted variables.
RESULTS: One hundred nine patients classified as high STarT risk had complete intake and discharge FS and DP data. Prevalence rate for DP was 65.1%. A significant and clinically important difference (7.98 FS points; p = .03) in change in function at discharge between DP and No-DP was observed after controlling for all confounding variables in the final model.
CONCLUSION: Findings suggest that interventions matched to directional preference (DP) are effective for managing high psychological risk patients and may provide physiotherapists with an alternative treatment pathway compared to managing similar patients with cognitive-behavioural approaches. Stricter research designs are required to validate study conclusions.
KEYWORDS: STarT; directional preference; low back pain; outcome measurement
From the FULL TEXT Article:
Directional preference (DP) is an important examination finding identified
within many treatment-based physiotherapy classification systems
in the United States and internationally (Delitto, Erhard, & Bowling,
1995; McKenzie & May, 2004; Petersen et al., 2003). DP is identified
if the patient's most distal pain intensity decreased, abolished, or
centralized, and/or their lumbar range of motion improved in response
to repeated end range movement tests or positional loading strategies
following McKenzie methods (McKenzie & May, 2004). The prevalence
rates reported for DP observed during initial evaluation ranged
between 60% and 78% (May & Aina, 2012). Clinical interest by physiotherapists
for identifying DP during the initial evaluation of patients
with non-specific low back pain (NSLBP) is supported by recent data
published in several reviews (Long, Donelson, & Fung, 2004; Long,
May, & Fung, 2008; Werneke et al., 2011).
Evidence also suggests that when DP is used as a treatment effect
modifier to inform treatment of patients with NSLBP, clinical outcomes
are improved (May & Aina, 2012). Specifically, clinical trials have
demonstrated significantly better treatment outcomes when patients
are randomized to the appropriate DP exercises compared to other
types of exercises or interventions (Brennan et al., 2006; Browder,
Childs, Cleland, & Fritz, 2007; Long et al., 2004). For example, Long
et al. (2004) randomized patients into treatment groups either
matching or unmatching exercises according to the patient's DP. The results demonstrated greater reductions in disability over a 2-week follow-up period when the specific exercise regimen was matched to the patient's DP compared to the group receiving unmatched exercise prescriptions. Long et al. (2008) also reported that patients with a DP who received matched exercise treatment according to the patient's DP had a 7.8 times greater likelihood of a good outcome compared with those who did not receive matched exercises (Long et al., 2008).
The spine literature recognizes that patients with NSLBP are biopsychosocially complex because of known interacting effects between sociodemographic, psychosocial, and biomechanical or physical factors experienced by each individual patient (O'Sullivan, Caneiro, O'Keeffe, & O'Sullivan, 2016). Psychosocial screening is considered a high priority for clinicians to identify patients for elevated pain-related psychosocial distress because such distress is a known precursor for increased risk of chronic disability and/or delayed treatment recovery (George & Beneciuk, 2015). To assist clinicians to accurately identify patients by psychosocial risk, Hill et al. (2008) developed a nine-item multidimensional biopsychosocial screening measure, that is, Subgroups for Targeted Treatment (STarT) Back Screening Tool. A primary purpose for STarT screening was to target treatment strategies for each of the three STarT risk patient subgroups, that is, low, medium, or high.
For patients with NSLBP who are classified at high psychosocial risk during the initial evaluation, contemporary clinical practice guidelines (Airaksinen et al., 2006) recommend managing these patients according to the biopsychosocial model, which endorses interdisciplinary treatments combining medical, physiotherapy, and mental health care services to improve patient outcomes and decrease health care costs (Hill et al., 2011; Foster et al., 2014). For those patients classified at high-risk using the STarT tool, the recommended treatment is formal cognitive-behavioural approaches (CBA) augmented by evidencebased physiotherapy intervention. To equip front-line clinicians with the necessary CBA intervention skills to effectively manage these high-risk patients, physicians and physiotherapists were recommended to attend advanced CBA educational and training programs (Sowden et al., 2012). Foster et al. (2014) reported large clinically important differences in 6-month patient self-report disability outcomes for high STarT risk patients prescribed with CBA-targeted treatments following STarT stratified management approach compared to similar patients receiving usual care in family practice.
There is scant guidance for physiotherapists treating patients with NSLBP when both favourable (i.e., DP) and unfavourable (i.e., high STarT risk) prognostic indicators are identified during the initial examination. The purpose of this study was therefore to examine the association between changes in functional status (FS) outcomes at discharge from rehabilitation among high STarT risk patients experiencing low back pain with a DP versus No-DP at baseline. We hypothesize that patients classified at high STarT risk whose symptoms show a first-visit DP and whose treatment was matched to their DP will achieve better FS outcomes at discharge from physiotherapy services compared to those high STarT risk patients whose symptoms do not demonstrate DP. Because prior evidence suggests that patients demonstrating DP can be rapidly and successfully treated with the performance of matching directional exercises (Long et al., 2004; Long et al., 2008), we believe that high STarT risk patients with DP are actually low-risk patients despite their baseline STarT Back score. If results from this study are promising, further research using stricter designs would be required to examine if targeted treatment using DP-matched exercises versus the STarT stratification treatment approach produces better outcomes.
Data were collected between January 2013 and May 2017. All participating
clinicians collected patient data and outcomes using an international
medical rehabilitation data management company, that is, Focus
On Therapeutic Outcomes, Inc. (FOTO), Knoxville, TN, USA (Swinkels
et al., 2007). The Rutgers University Institutional Review Board
approved the project. The study did not include changes in clinical
practice, data documentation, or treatment; therefore, patient
informed consent was not required.
One hundred thirty-eight patients older than 17 years of age
experiencing NSLBP complaints, classified into a high STarT risk category
at baseline, and referred to a participating outpatient physiotherapy
clinic were included. Patients were excluded due to pregnancy or
suspicion of serious spinal pathology such as fracture, cancer, or visceral
Eight physiotherapists, credentialed in McKenzie methods (McKenzie
& May, 2004), participated in data collection. Four clinicians achieved
the highest level of McKenzie training and were credentialed by the
International McKenzie Institute as Diplomats; and four clinicians were
certified for achieving basic competency in McKenzie methods. Practice
settings were diverse across private practice, hospital-based, and
military outpatient physiotherapy clinics. Although participating physiotherapists
were not blinded to STarT risk classification, no physiotherapist
in this study received formal CBA training as recommended
by the STarT management approach (Hill et al., 2008; Sowden et al.,
Patients completed the STarT and FOTO's FS Lumbar Computer AdaptiveTest
(LCAT) measures at intake (Hart, Werneke, Wang, Stratford, &
Mioduski, 2010; Hill et al., 2008). The LCAT is a psychometrically
strong outcome measure to assess the patient's FS and change in FS
(Hart et al., 2010; Hart, Mioduski, Werneke, & Stratford, 2006). The
LCAT FS measure was repeated at discharge. In addition, the following
intake patient and physiotherapist characteristics were included in the
model, because they are shown to influence FS outcomes: age, gender,
symptom duration, payer, use of medication at intake, exercise history,
lumbar surgical history, prior treatment, condition complexity determined
29 possible comorbid conditions, McKenzie credentialing level,
and treatment duration (Table 1).
STarT Back Screening Tool stratification and scoring techniques have
been described in detail elsewhere (Hill et al., 2008). Research findings
have supported good reliability, validity, clinical utility, and usefulness
for screening patients with lumbar impairments using STarT Back
Screening Tool (Hill et al., 2011; Foster et al., 2014; Von Korff et al.,
2014; Wideman et al., 2012).
The operational definition for examining DP used in our study has
been previously described and recommended (Long et al., 2008;
Werneke et al., 2011). DP was identified if the patient's most distal
pain intensity decreased, abolished, or centralized, and/or their lumbar
range of motion improved in response to repeated end range movement
tests or positional loading strategies. Interrater chance corrected
agreement (Kappa) for identifying a DP for patients' whose symptoms
centralized has been reported to be excellent (Kappa = .90; Kilpikoski
et al., 2002).
Treatment processes were guided by DP, if present. For example, if
lumbar extension was identified as the patient's DP, exercises that
moved the patient towards lumbar extension and manual procedures
that produced lumbar extension forces were applied. If patients were
classified into a No-DP category, an individualized active rehabilitation
plan was determined at the discretion of the treating therapist. All
patients regardless of DP received the same educational approach:
Empower the patient to become actively involved in his or her own
recovery to reduce fear of physical activity and movement intolerance.
There was no attempt to standardize care beyond these guidelines.
FOTO's LCAT FS measure used in the study has been previously
described in detail (Hart et al., 2006; Hart et al., 2010). Briefly, FS
scores estimated by the LCAT FS measure ranged from 0 (low) to
100 (high functioning) on a linear metric. The items in the LCAT FS measure
item bank have demonstrated internal consistency, reliability,
validity, sensitivity, and responsiveness (minimal clinically important
difference averaged 5 points).
The percentage (%) of patients classified as first-visit DP or No-DP was
calculated. Patient and therapist characteristics are reported inTable 1.
Univariate and multivariate regression models examined the association
between DP and No-DP patient subgroups and change in FS at
discharge from rehabilitation services.
Our multivariate model evaluated for confounding by the following
11 risk-adjusted variables:
medication use at intake,
prior treatment episodes,
treatment duration, and
post graduate McKenzie education and training.
Confounding effects were determined
using the criteria that a 10% change in the effect measure when
adding the potential confounder(s) to the baseline model represents
meaningful confounding (SAS Institute Inc., 1999). We calculated beta
coefficients for DP versus No-DP subgroups in which the reference
standard was DP. Variables entering the model were checked for collinearity
by evaluating the condition index.
We examined the potential for bias due to loss to follow-up by
comparing patient and therapist characteristics for those patients with
complete FS and DP data at intake versus those with missing FS data
at discharge (Table 2).
One-hundred and nine subjects classified as high STarT risk at baseline
had DP intake and discharge FS data. Of these, 65.1% were classified
as having DP versus 34.9% having No-DP. Twenty-nine patients were
lost to follow up. The completion rate (i.e., complete intake and discharge
outcome data) was 79%. There were no differences between
drop-outs and those who completed treatment in 7 of the 11 model
variables (Table 2) including intake FS, which was the strongest variable
in the model. However patients who dropped-out were more
likely to have No-DP, chronic symptoms, use of pain medications,
and longer treatment duration.
Overall, subjects' FS improved from 42.2 to 69.6 FS points, for a
total increase of 27.4 units. Scores among subjects who demonstrated
a DP at intake increased from 43.6 to 74.4, a change of 30.8 points
over an average of 7.5 visits, whereas those who did not demonstrate
a DP at intake increased from 39.5 to 60.6, a change of 21.0 points
while using an average of 8.0 visits. These findings represent a significant
association in change in FS favouring high STarT risk patients who
demonstrated a DP (p = .0009, beta coefficient [–]10.82) versus No-
We evaluated for potential confounding by the aforementioned
11 risk-adjustment variables.
The following seven variables demonstrated
meaningful confounding effects:
duration of symptoms,
prior treatment episode,
condition complexity, and
post graduate McKenzie education.
The condition index for this
analysis was 45.9. Analyses with condition indexes above 30 are considered
to be significantly affected by colinearity (SAS Institute Inc., 1999). Subsequent analyses showed that age and condition complexity
demonstrated colinearity effects in this association and was therefore
deleted from the multivariate model. With age and condition complexity
deleted from the model, the condition index decreased to 29.5. This
model was rechecked for meaningful confounding, and all remaining
variables were retained.
The final model therefore included
intake FS score,
duration of symptoms,
prior treatment episode, and
post graduate McKenzie education.
In the final model,
there was a significant and clinically important difference, that is,
(–)7.98 (p = .03), in change in FS at discharge between DP and No-DP.
The purpose of this study was to examine the association between
changes in FS at discharge from rehabilitation for high STarT risk
patients experiencing low back pain whose symptoms showed a DP
versus No-DP at baseline. First-visit DP was both significant and
clinically important for explaining FS outcomes at discharge. We had
anticipated that DP data may be predictive of good FS outcomes, in
part because the treatment prescribed by the participating therapists
in our study was matched to the patient's DP. Evidence suggests that
DP is associated with favourable treatment outcomes when patients
were randomized to the appropriate DP-matched exercises (Brennan
et al., 2006; Browder et al., 2007; Long et al., 2004).
There are scant data examining DP as a baseline prognostic variable
for predicting FS outcomes (Delitto et al., 2012; May & Aina,
2012). In the only study that we are aware of which examined DP's
prognostic value, the authors reported that first-visit DP was predictive
of pain intensity but not FS at discharge for patients with low back
pain referred to physiotherapy services (Werneke et al., 2011). In contrast
with this earlier study, our results indicated that DP was associated
with improved FS outcomes. Differences in sample sizes, patient
characteristics including high STarT risk, and number of independent
variables controlled for in the models could partially explain contrasting
FS outcome results between our current study and previous findings
(Werneke et al., 2011).
Our inclusion criteria for this study were restricted to patients
with NSLBP who were classified at high STarT psychosocial risk. Prior
research has shown that these patients have a particularly poor treatment
prognosis if not managed by comprehensive multifactorial interventions
as demonstrated by the STarT approach combining a formal
CBA with evidence-based physiotherapy interventions (Hill et al.,
2008). The STarT stratification approach applying prognostic screening
to matched treatment pathways is recommended for improved shortand
long-term health benefits, improved function, and cost savings
compared to usual care (Hill et al., 2011; Foster et al., 2014).
Despite recommendations for targeting patients classified at high
STarT risk for formal CBA interventions, the majority of high STarT risk
patients in our study not only demonstrated DP during the initial evaluation
but also responded well with significant and clinically important
increases in FS scores at discharge compared to the No-DP subgroup.
Patients at high risk who demonstrated DP were not managed by
STarT stratification treatments but were prescribed specific exercises
matched to DP based on McKenzie methods. Our data suggest that
physiotherapists credentialed in McKenzie methods can consider DP-matched exercises as a potential alternative method to STarT stratification
treatments to improve functional outcomes when managing
high STarT risk patients with NSLBP.
The foundation of the McKenzie method is built on patient education
consisting of empowerment, reassurance, and active self-management
encouraging patients to adhere to their intervention plan during
therapy sessions and at home between treatment sessions (May,
2007). Many of these patient self-care principles are similar to the
patient educational strategies emphasized and reinforced within formal
cognitive behavioural training programs (Main, Sowden, Hill, Watson,
& Hay, 2012; Sowden et al., 2012). Evidence suggests positive
effects of the McKenzie system for decreasing individual psychosocial
risk factors, such as elevated distress or elevated fear avoidance
beliefs, while simultaneously improving patients' functional outcomes
(Al-Obaidi, Al-Sayegh, Ben Nakhi, & Skaria, 2013, Werneke et al.,
It is therefore possible that our results could have been
associated with these McKenzie biopsychosocial patient educational
principles utilized by clinicians trained in the McKenzie approach independent
of DP or No-DP. However, the same McKenzie patient educational
approach was utilized regardless of DP classification category.
Although both groups, that is, DP and No-DP, improved FS during
the episode of care, if the educational approach was solely responsible
for the observed association between DP and FS outcomes, we would
have expected no significant difference between DP and No-DP subgroup
outcomes. It is therefore plausible that the presence of a DP
enhances the effects of McKenzie patient education strategies. Nevertheless,
we had no hypotheses regarding how high STarT risk stratification
would interact with specific physical therapy interventions,
because causality between interventions and outcome was not possible
based on the study design.
Participating physiotherapists in our study as well as others
(Beneciuk et al., 2013; Beneciuk, Fritz, & George, 2014; Fritz,
Beneciuk, & George, 2011) did not receive formal CBA training and
did not follow recommended STarT-targeted treatment pathways.
Despite physiotherapists' lack of formal CBA education and training,
significant improvements in FS scores at discharge from physiotherapy
for high-risk STarT patients with NSLBP have been consistently
reported (Beneciuk et al., 2013, 2014; Fritz et al., 2011). Participating
therapists in these other studies were trained in a popular physiotherapy
treatment-based classification approach that included DP in its
algorithm (Delitto et al., 1993). So, despite initial favourable results
supporting the importance of formal CBA training to improve patient
outcomes and reduce health care costs (Hill et al., 2011, Foster et al.,
2014, von Korff et al., 2014, Wideman et al., 2012), preliminary evidence
from our study suggests that DP may be an effective alternative
approach for managing patients with elevated pain-related psychosocial
distress and that formal CBA training may not be required to
achieve successful patient outcomes.
Not all patients classified at high STarT risk in our study (21%) completed
physical therapy care; therefore, our results need to be
interpreted in light of missing data. However, our completion rate
was only slightly lower compared to other studies examining physiotherapy
outcomes related to STarT stratification. For example, drop-out rates of approximately 18% and 16% were reported respectively
by Fritz et al. (2011) and Beneciuk et al. (2014). However, these two
studies reported drop-out rates for the entire sample regardless of
baseline STarT risk levels, whereas our drop out results pertained specifically
for patients with higher biopsychosocial complexity who were
classified into the high STarT risk category.
Our study design was a case series examining a sample of convenience,
which limits the generalizability of our results. In addition, we
cannot infer a cause and effect between any treatments and outcomes
observed in our study nor suggest that treatments based on DP are
more effective compared to STarT-recommended treatments for
patients classified at elevated psychosocial risk. Stronger research
designs are required to validate our study's findings.
Regardless of best efforts to adjust for differences in our patient
case mix, we acknowledge that a limitation of observational research
is that we can only control for potential confounders that we measured.
There may be unmeasured variables over which we have no
control or simply did not collect data on that might have influenced
our study results, for instance, a delay in treatment time, administration
policy regarding the number of patients treated by the therapist
per hour, or other patient socio-economic status and financial
resources to attend physiotherapy services.
Study results suggest that patients who are at high STarT risk and demonstrate
DP during the first visit report greater improvements in function
than those high STarT risk patients who do not demonstrate firstvisit
DP. These findings suggest that DP may provide an alternative
treatment pathway compared to targeted interventions recommended
by STarT proponents (Hill et al., 2008; Foster et al., 2014, Sowden
et al., 2012). Nevertheless, future research using stricter designs is
warranted to validate our results.
All authors have contributed to either the data collection, writing and
preparing the manuscript, or statistical analyses of the data. All authors
have given final approval for the manuscript submitted to Physiotherapy
Airaksinen, O., Brox, J. I., Cedraschi, C., Hildebrandt, J., Klaber-Moffett, J. (2006).
Chapter 4. European guidelines for the management of chronic nonspecific low back pain
European Spine Journal, 15, S192–S300.
Al-Obaidi, S. M., Al-Sayegh, N. A., Ben Nakhi, H., & Skaria, N. (2013).
Effectiveness of McKenzie intervention in chronic low back pain:
A comparison based on the centralization phenomenon utilizing selected
bio-behavioral and physical measures.
International Journal of Physical Medicine Rehabilitation, 1, 4.
Beneciuk, J. M., Bishop, M. D., Fritz, J. M., Robinson, M. E., Asal, N. R. (2013).
The STarT Back Screening Tool and individual psychological measures:
Evaluation of prognostic capabilities for low back pain clinical outcomes
in outpatient physical therapy settings.
Physical Therapy, 93, 321–333.
Beneciuk, J. M., Fritz, J. M., & George, S. Z. (2014).
The STarT Back Screening Tool for prediction of 6-month clinical outcomes:
Relevance of change patterns in outpatient physical therapy settings.
Journal of Orthopaedic & Sports Physical Therapy, 44, 656–664.
Brennan, G. P., Fritz, J. M., Hunter, S. J., Thackeray, A., Delitto, A. (2006).
Identifying subgroups of patients with acute/subacute “nonspecific” low back pain:
Results of a randomized clinical trial.
Spine, 31, 623–631.
Browder, D. A., Childs, J. D., Cleland, J. A., & Fritz, J. M. (2007).
Effectiveness of an extension-oriented treatment approach in a subgroup of subjects
with low back pain: A randomized clinical trial.
Physical Therapy, 87, 1608–1618.
Delitto, A., Erhard, R. E., & Bowling, R. W. (1995).
A treatment-based classification approach to low back syndrome:
Identifying and staging patients for conservative treatment.
Physical Therapy, 75, 470–485.
Delitto, A., George, S. Z., Van Dillion, L., Whitman, J. M., Sowa, G. (2012).
Low back pain: Clinical practice guidelines linked to the International Classification of
Functioning, Disability, and Health from the orthopaedic section of the American Physical
Journal of Orthopaedic & Sports Physical Therapy, 42, A1–A57
Foster, N. E., Mullis, R., Hill, J. C., Lewis, M., Whitehurst, D., Doyle, C. (2014).
Effect of stratified care for low back pain in family practice (IMPaCT back):
A prospective population-based sequential comparison.
Annals of Family Medicine, 12, 102–111.
Fritz, J. M., Beneciuk, J. M., & George, S. Z. (2011).
Relationship between categorization with the STarT Back Screening Tool and prognosis
for people receiving physical therapy for low back pain.
Physical Therapy, 91, 722–732.
George, S. Z., & Beneciuk, J. M. (2015).
Psychological predictors of recovery from low back pain: A prospective study.
BMC Musculoskeletal Disorders, 16, 49
Hart, D. L., Mioduski, J. E., Werneke, M. W., & Stratford, P. W. (2006).
Simulated computerized adaptive test for patients with lumbar spine impairments
was efficient and produced valid measures of function.
Journal of Clinical Epidemiology, 59, 947–956.
Hart, D. L., Werneke, M. W., Wang, Y. C., Stratford, P. W., & Mioduski, J. E.
Computerized adaptive test for patients with lumbar spine impairments produced valid
and responsive measures of function.
Spine, 35, 2157–2164.
Hill, J. C., Dunn, K. M., Lewis, M., Mullis, R., Main, C. J., & Foster, N. E.
A primary care back pain screening tool:
Identifying patient subgroups for initial treatment.
Arthritis and Rheumatism, 59, 632–641.
Hill, J. C., Whitehurst, D. G., Lewis, M., Bryan, S., Dunn, K. M., & Foster, N.
Comparison of stratified primary care management for low back pain with current best
practice (STarT Back): A randomised controlled trial.
Lancet, 378, 1560–1571.
Kilpikoski, S., Airaksinen, O., Kankaanpa, M., Leminen, P., Videman, T., &
Alen, M. (2002).
Interexaminer reliability of low back pain assessment using the McKenzie method.
Spine, 27, E207–E214.
Long, A., Donelson, R., & Fung, T. (2004).
Does it matter which exercise? A randomized control trial of exercise for low back pain.
Spine, 29, 2593–2602.
Long, A., May, S., & Fung, T. (2008).
The comparative prognostic value of directional preference and centralization:
A useful tool for front-line clinicians?
The Journal of Manual & Manipulative Therapy, 16, 248–254.
Main, C. J., Sowden, G., Hill, J. C., Watson, P. J., & Hay, E. M. (2012).
Integrating physical and psychological approaches to treatment in low back pain:
The development and content of the STarT Back trial's ‘high-risk’ intervention
(StarT Back; ISRCTN 37113406).
Physiotherapy, 98, 110–116.
May, S. (2007).
Patients' attitudes and beliefs about back pain and its management after physiotherapy
for low back pain.
Physiotherapy Research International, 12, 126–135.
May, S., & Aina, A. (2012).
Centralization and directional preference: A systematic review.
Manual Therapy, 17, 497–506.
McKenzie, R., & May, S. (2004).
The lumbar spine: Mechanical diagnosis and therapy. 2nd ed.
Waikanae, New Zealand: Spinal Publications; 2003.
Phenomenon of spinal symptoms—A systematic review.
Manual Therapy, 9, 134–143.
O'Sullivan, P., Caneiro, J. P., O'Keeffe, M., & O'Sullivan, K. (2016).
Unraveling the complexity of low back pain.
Journal of Orthopaedic & Sports Physical Therapy, 46, 932–937
Petersen, T., Laslett, M., Thorsen, H., Manniche, C., Ekdahl, C., & Jacobsen,
Diagnostic classification of non-specific low back pain.
A new system integrating patho-anatomic and clinical categories.
Physiotherapy Theory and Practice, 19, 213–237.
SAS Institute Inc (1999).
SAS OnlineDoc®, Version 8.
Cary, NC: SAS Institute Inc.
Sowden, G., Hill, J. C., Konstantinou, K., Meenee Khannaa, D., Chris, J. (2012).
IMPaCT Back study team. Targeted treatment in primary care for low back pain:
The treatment system and clinical training programs used in the IMPaCT Back study.
Family Practice, 29, 50–62.
Swinkels, I. C., van den Ende, C. H., de Bakker, D., van der Wees, P., Hart, D. (2007).
Clinical databases in physical therapy.
Physiotherapy Theory and Practice, 23, 153–167.
Von Korff, M., Shortreed, S. M., Saunders, K. W., Leresche, L., Berlin, J. A. (2014).
Comparison of back pain prognostic risk stratification item sets.
The Journal of Pain, 15, 81–89.
Werneke, M. W., Hart, D. L., Cutrone, G., Oliver, D., McGill, T., Weinberg, J. (2011).
Association between directional preference and centralization in patients with low back pain.
Journal of Orthopaedic & Sports Physical Therapy, 41, 22–31
Werneke, M. W., Hart, D. L., George, S. Z., Stratford, P. W., Matheson, J. W. (2009).
Clinical outcomes for patients classified by fear-avoidance beliefs
and centralization phenomenon.
Archives of Physical Medicine and Rehabilitation, 90, 768–777.
Wideman, T. H., Hill, J. C., Main, C. J., Lewis, M., Sullivan, M. J. (2012).
Comparing the responsiveness of a brief, multidimensional risk screening tool for back pain
to its unidimensional reference standards: The whole is greater than the sum of its parts.
Pain, 153, 2182–2191.
Return to the LOW BACK PAIN Section