ORIGINAL ARTICLE
Does the addition of visceral manipulation alter outcomes for patients with low back pain? A randomized placebo controlled trial J. Panagopoulos1, M.J. Hancock1, P. Ferreira2, J. Hush1, P. Petocz3 1 Faculty of Human Sciences, Macquarie University, Sydney, Australia 2 Faculty of Health Sciences, University of Sydney, Australia 3 Department of Statistics, Macquarie University, Sydney, Australia
Correspondence John Panagopoulos E-mail: [email protected] Funding sources None. Conflicts of interest None declared. Accepted for publication 23 September 2014 doi:10.1002/ejp.614
Abstract Background: This study aimed to investigate whether the addition of visceral manipulation, to a standard physiotherapy algorithm, improved outcomes in patients with low back pain. Methods: Sixty-four patients with low back pain who presented for treatment at a private physiotherapy clinic were randomized to one of two groups: standard physiotherapy plus visceral manipulation (n = 32) or standard physiotherapy plus placebo visceral manipulation (n = 32). The primary outcome was pain (measured with the 0–10 Numerical Pain Rating Scale) at 6 weeks. Secondary outcomes were pain at 2 and 52 weeks, disability (measured with the Roland-Morris Disability Questionnaire) at 2, 6 and 52 weeks and function (measured with the Patient-Specific Functional Scale) at 2, 6 and 52 weeks. This trial was registered with the Australia and New Zealand Clinical Trials Registry (ACTRN12611000757910). Results: The addition of visceral manipulation did not affect the primary outcome of pain at 6 weeks (−0.12, 95% CI = −1.45 to 1.21). There were no significant between-group differences for the secondary outcomes of pain at 2 weeks or disability and function at 2, 6 or 52 weeks. The group receiving addition of visceral manipulation had less pain than the placebo group at 52 weeks (mean 1.57, 95% CI = 0.32 to 2.82). Participants were adequately blinded to group status and there were no adverse effects reported in either group. Conclusions: Our study suggests that visceral manipulation in addition to standard care is not effective in changing short-term outcomes but may produce clinically worthwhile improvements in pain at 1 year.
1. Background Despite the large number of randomized controlled trials investigating interventions for low back pain (LBP), recommended interventions produce only small effects (Keller et al., 2007). It is possible that outcomes could be improved by combining current interventions or by the addition of new therapies. Visceral manipulation is a manual therapy technique aimed at treating motion abnormalities of the internal organs (Barral, 2005; Vleminckx, 2006). © 2014 European Pain Federation - EFIC®
Although visceral manipulation is not taught in basic physiotherapy training, training in this treatment technique is being sought out by many physiotherapists and health practitioners in post-graduate workshops. However, there is paucity of research into the efficacy of this approach. There is preliminary evidence that visceral manipulation may be effective for treating LBP. A recent clinical series demonstrated improved symptoms following a specific visceral manipulation technique aimed at mobilizing the kidneys in people with nonspecific LBP Eur J Pain •• (2014) ••–••
1
Does visceral manipulation alter low back pain outcomes?
What’s already known about this topic? • Visceral manipulation has been used for over a decade in the treatment of low back pain, but its efficacy remains untested. What does this study add? • This study demonstrates that the addition of visceral manipulation to a standard physiotherapy treatment algorithm improves long-term but not short-term pain.
(Tozzi et al., 2012). Another within-subject, repeated measures study demonstrated immediate lumbar pressure/pain threshold reduction in asymptomatic subjects following visceral manipulation techniques (McSweeney et al., 2012). The rationale for this therapeutic approach is that visceral disorders could potentially trigger or exacerbate LBP symptoms in the presence of impaired movement between internal organs and their supportive tissues. This could manifest as LBP via two possible mechanisms: visceral referred pain and central sensitization. The mechanism by which visceral pain causes referral to somatic structures could occur by neural convergence, whereby sympathetic afferent nerves that convey signals from the viscera converge with somatic nerves in the dorsal horn. Due to the low proportion of visceral to somatic afferents entering the dorsal horn (Cervero, 2000) and because activation of certain visceral receptors does not induce conscious perception (Cervero, 2009), visceral nociceptive input can be misinterpreted as arising from somatic structures. Activation of visceral nociceptors can result in central sensitization. These nociceptors can be stimulated by an altered gut environment or by altered gastrointestinal/urinary motility (Cervero, 2009). Hence, small variations around the visceral sensory receptors can evoke peripheral hypersensitivity (Cervero, 1995). Even a low level of peripheral input can cause increased activity of projection neurons in the spinal cord and can dynamically maintain this central activation (Woolf, 2011). Central sensitization can result in normal sensory stimuli, such as mechanical touch, being experienced as pain (Woolf, 2011). The mechanisms by which visceral manipulation may have an effect on pain are not yet understood. One hypothesis is that by specific manual treatment of the supportive fascia of the internal cavities of the trunk, visceral manipulation modulates visceral nociceptive signalling (Vleminckx, 2006), reducing excessive visceral nociceptive input entering the spinal cord 2 Eur J Pain •• (2014) ••–••
J. Panagopoulos et al.
and allowing hypersensitive central neurons to return to a normal state of excitation (Woolf, 2011). Therefore, it is possible that visceral disorders contribute to the development and continuation of LBP in some patients and that standard therapies which do not directly target the viscera (and surrounding fascia) could neglect a potential pain-producing component. The efficacy of visceral manipulation for LBP has not yet been evaluated in a randomized controlled trial. The aim of this study was to investigate whether the addition of visceral manipulation to standard physiotherapy care for LBP improves pain, disability and functional outcomes.
2. Methods 2.1 Study design The study design was a randomized placebo controlled trial with blinded follow-up. Ethical approval for this study was received from the University of Sydney Human Research Ethics Committee. This study was registered with the Australia and New Zealand Clinical Trials Registry (ACTRN12611000757910) and the study protocol was published (Panagopoulos et al., 2013).
2.2 Study population/recruitment Consecutive patients who presented with LBP to a private physiotherapy clinic in Sydney, Australia were screened for eligibility. Potential participants who met the inclusion criteria were invited to participate in the trial and provided written consent. To be eligible for the trial, participants were required to meet all of the following criteria as assessed by the treating physiotherapist: primary complaint of pain in the area extending from the 12th rib to the buttock crease (this may or may not be accompanied by pain in the leg or other spinal areas); LBP symptoms which have a score of >2/10 on the 0–10 Numerical Pain Rating Scale (NPRS) (Pengel et al., 2004); aged 18–80; no known or suspected serious spinal pathology (e.g., metastatic, inflammatory or infective diseases of the spine, cauda equina syndrome, canal stenosis, spinal fracture); no nerve root compromise evidenced by at least two of the following (1) myotomal weakness, (2) dermatomal or widespread sensory loss, (3) hypoor hyperreflexia of the lower limb reflexes; no spinal surgery within the preceding 6 months; no visceral surgery within the preceding 6 months; no vascular abnormality such as abdominal aortic aneurysms; not currently receiving chiropractic, osteopathic or other physical therapy; not pregnant or suspect being pregnant; not currently in an acute inflammatory phase of known gastrointestinal or urinary diseases (such as cholecystitis, renal calculi, peritonitis, appendicitis); not currently taking medications that significantly alter gut motility; not currently taking medications (such as oral cor-
© 2014 European Pain Federation - EFIC®
J. Panagopoulos et al.
Does visceral manipulation alter low back pain outcomes?
ticosteroids which are known to increase the risk of intestinal perforation); no known gastrointestinal disease that associates with a risk of intestinal perforation (e.g. Crohn’s disease, diverticular disease, peptic ulcer disease); not taking anti-platelet medications such as warfarin.
2.3 Baseline measures and randomization Participants completed baseline data immediately prior to being randomized to a treatment arm of the study (Fig. 1). A total of 98 patients were eligible and 34 were excluded as they did not fit the inclusion criteria (Fig. 1). Of these excluded patients, 18 chose not to participate, 12 had a pain rating which was too low at baseline, two were pregnant, one had had previous lumbar surgery and one patient was over 80 years of age (Fig. 1). At initial assessment, demographic data and the following baseline measures were collected: pain, measured with the 0–10 NPRS (Pengel et al., 2004), where 0 = no pain and 10 = pain as bad as it could be; disability, measured with the 0–24 RolandMorris Disability Scale (Roland and Morris, 1983), where 0 = no disability and 24 = severe disability; function, measured with the 0–10 Patient-Specific Functional Scale (Stratford et al., 1995), in which the patient nominates three important activities which are
limited by their pain and rates their ability to perform them on a Likert scale ranging from 0 (unable to perform the activity) to 10 (able to perform the activity at pre-injury level). The scores are summed and averaged producing a score out of 10; presence of gastrointestinal/urinary/reproductive symptoms was assessed by asking participants if they suffer from bloating, diarrhoea, constipation, period pain, cramping, food sensitivity or reflux. This variable of visceral symptomatology allowed researchers to determine whether patients with obvious visceral symptoms were randomized equally to both treatment arms of the study. A researcher not involved in data collection or analysis developed a randomization schedule using Excel to generate 64 sequentially numbers, which were concealed in sealed, opaque randomization envelopes. These envelopes contained a paper with words ‘VM’ or ‘placebo’. After baseline data had been collected, the treating physiotherapist opened the next randomization envelope and allocated the participant according to the randomization schedule (Fig. 1).
2.4 Clinical assessment Participants were assessed by one of two experienced musculoskeletal physiotherapists. Both physiothera-
98 Eligible LBP patients attend for physiotherapy assessment and treatment
34 patients excluded 18 chose not to participate 12 pain rating too low 2 pregnant 1 previous lumbar surgery
64 participants randomized
1 >80 years of age
32 Placebo group
32 Visceral manipulation group
standard physiotherapy + sham visceral manipulation
standard physiotherapy + real visceral manipulaion
31 followed-up at 2 weeks (97%)
32 followed-up at 2 weeks (100%)
30 followed-up at 6 weeks (94%)
32 followed-up at 6 weeks (100%)
28 followed-up at 52 weeks (88%)
31 followed-up at 52 weeks (97%)
Figure 1 Study flow diagram.
© 2014 European Pain Federation - EFIC®
Eur J Pain •• (2014) ••–••
3
Does visceral manipulation alter low back pain outcomes?
pists had concluded post-graduate intensive practical and coursework training in visceral manipulation and had between 6 and 10 years of clinical experience in the use of visceral manipulation. All participants received a standardized assessment involving active movement testing, passive intervertebral motion testing, palpation, neural tension testing and functional stability testing. Participants also received palpation testing of their visceral system to assess which visceral structures may be involved in the participant’s clinical presentation. This palpation testing involved the gastrointestinal, urinary, respiratory, reproductive and cardiovascular systems as described in detail in Barral (2005). In brief, the physiotherapist placed his hands on the anterior part of the subject’s body and assessed organ motility and visceral ligament mobility. For participants allocated to placebo visceral manipulation, these findings were noted but not used in treatment.
2.5 Treatments Participants in both groups received the same standardized physical examination and standard care. The two treating physiotherapists treated both control and intervention patients. As per LBP guidelines, all participants received current evidence-based advice focusing on remaining active (Koes et al., 2010). Participants received manual therapy, exercise therapy (lumbar muscle re-training and functional exercise prescription) and massage, as required by the treating physiotherapist. To facilitate lumbar muscle re-training, real-time ultrasound guided imaging was used to provide visual feedback and education. All participants were treated one to two times per week for a minimum of one week and a maximum of 12 treatments over 6 weeks. The number of treatment sessions was based on the rate of participants’ symptom reduction. Following the 6-week period, treatment was continued if participants felt their treatment or rehabilitation goals had not been met. Participants continued to receive treatment as per their group allocation and remained blinded. For both groups, initial treatment sessions lasted for approximately 40 min and follow-up sessions lasted approximately 25–30 min. In order to ensure compliance and monitor treatments delivered in the study, therapists kept a record of the number of times the patient attended physiotherapy and recorded details of the treatment including the techniques used. The interventions for each group were as follows: 4 Eur J Pain •• (2014) ••–••
J. Panagopoulos et al.
2.5.1 Standard care plus active visceral manipulation group In addition to standard care (as described above), any fascial restrictions were treated using specific visceral manipulation techniques (Barral, 2005; Tozzi et al., 2012). This took approximately 5–10 min and involved light or deep manual fascial releases and specific organ mobilizations in the thoracic, subdiaphragmatic, abdominal and pelvic areas as appropriate (Barral, 2005). 2.5.2 Standard care plus placebo visceral manipulation group Participants were treated with standard care as defined above. Participants also received a placebo visceral ‘treatment’ which involved approximately 5 min of sham treatment aimed around the abdominal area. This involved light touch and no intention on the part of the physiotherapist to impart any therapeutic technique. This placebo therapy was performed on areas of the abdomen not involved in a mechanical, functional or neurological sense to any visceral issues present. The placebo technique was pilot tested on experienced physiotherapists (who had no prior experience of visceral manipulation) prior to starting the trial. The physiotherapists were unable to distinguish between the placebo and real visceral manipulation.
2.6 Outcome measures Measures of outcome were recorded by an assessor blinded to group allocation. Outcomes were recorded at baseline, 2, 6 and 52 weeks after commencement of treatment. If the participant had ceased treatment, the outcome measures were collected by a blinded assessor over the phone or by email. At 6 weeks, participants were asked a treatment credibility question regarding which additional treatment they thought they received (real or placebo treatment). This was performed to assess if participant blinding was successful. 2.6.1 Primary outcome The primary outcome was pain intensity (0–10 NPRS) at 6 weeks as it was hypothesized to be the time when the intervention may have largest effect. 2.6.2 Secondary outcomes The secondary outcomes were pain intensity (0–10 NRPS) at 2 and 52 weeks, function and disability at 2, 6 and 52 weeks. © 2014 European Pain Federation - EFIC®
J. Panagopoulos et al.
Does visceral manipulation alter low back pain outcomes?
Table 1 Effects of visceral manipulation compared with control.
Visceral manipulation
Control
Adjusted treatment differences (95% confidence interval)
3.06 (2.08) 2.31 (1.99) 1.52 (1.65)
3.74 (2.25) 2.33 (2.22) 3.21 (2.27)
0.55 (−0.65 to 1.75) −0.12 (−1.45 to 1.21) 1.57 (0.32 to 2.82)
0.362 0.858 0.015
5.78 (5.40) 3.00 (2.96) 2.06 (3.56)
6.26 (5.35) 3.10 (3.98) 3.50 (3.61)
−0.86 (−3.33 to 1.58) −1.20 (−4.18 to 1.75) 0.11 (−2.86 to 3.07)
0.479 0.418 0.942
6.10 (2.13) 7.70 (1.81) 8.43 (1.76)
6.15 (1.95) 7.51 (1.86) 7.55 (1.82)
0.81 (−0.33 to 1.94) 0.61 (−0.63 to 1.84) −0.08 (−1.18 to 1.02)
0.158 0.332 0.882
Unadjusted mean outcomes (standard deviation) Outcome Pain (NPRS) 2/52 6/52 52/52 Disability (RMDQ) 2/52 6/52 52/52 Function (PSFS) 2/52 6/52 52/52
p-value
Numerical Pain Rating Scale (NPRS): where 0 = no pain and 10 = pain as bad as it could be. Roland-Morris Disability Questionnaire (RMDQ): where 0 = no disability and 24 = severe disability. Patient-Specific Functional Scale (PSFS): patient nominated three important activities which were limited by their injury and rated these on a Likert scale ranging from 0 (unable to perform the activity) to 10 (able to perform the activity at pre-injury level). The scores were summed and averaged producing a score out of 10.
2.7 Data analysis
3. Results
The sample size of 64 was determined a priori (Panagopoulos et al., 2013). This sample size was calculated to provide 80% power to detect a 1.5 point difference (which we considered to be a worthwhile effect) on the 0–10 NPRS, allowing for an estimated standard deviation of 2 and loss to follow-up of 10%. An ‘unstructured’ covariance matrix was assumed, although the results were very similar using a range of other structures (autoregressive, Toeplitz, HuynhFeldt, compound symmetry). All data were double entered. Data analysis was performed by a statistician who was blinded to group status and following intention-to-treat principles. The number of analyses was limited to reduce the possibility of type I errors. For primary outcomes, a p-value of
Does the addition of visceral manipulation alter outcomes for patients with low back pain? A randomized placebo controlled trial J. Panagopoulos1, M.J. Hancock1, P. Ferreira2, J. Hush1, P. Petocz3 1 Faculty of Human Sciences, Macquarie University, Sydney, Australia 2 Faculty of Health Sciences, University of Sydney, Australia 3 Department of Statistics, Macquarie University, Sydney, Australia
Correspondence John Panagopoulos E-mail: [email protected] Funding sources None. Conflicts of interest None declared. Accepted for publication 23 September 2014 doi:10.1002/ejp.614
Abstract Background: This study aimed to investigate whether the addition of visceral manipulation, to a standard physiotherapy algorithm, improved outcomes in patients with low back pain. Methods: Sixty-four patients with low back pain who presented for treatment at a private physiotherapy clinic were randomized to one of two groups: standard physiotherapy plus visceral manipulation (n = 32) or standard physiotherapy plus placebo visceral manipulation (n = 32). The primary outcome was pain (measured with the 0–10 Numerical Pain Rating Scale) at 6 weeks. Secondary outcomes were pain at 2 and 52 weeks, disability (measured with the Roland-Morris Disability Questionnaire) at 2, 6 and 52 weeks and function (measured with the Patient-Specific Functional Scale) at 2, 6 and 52 weeks. This trial was registered with the Australia and New Zealand Clinical Trials Registry (ACTRN12611000757910). Results: The addition of visceral manipulation did not affect the primary outcome of pain at 6 weeks (−0.12, 95% CI = −1.45 to 1.21). There were no significant between-group differences for the secondary outcomes of pain at 2 weeks or disability and function at 2, 6 or 52 weeks. The group receiving addition of visceral manipulation had less pain than the placebo group at 52 weeks (mean 1.57, 95% CI = 0.32 to 2.82). Participants were adequately blinded to group status and there were no adverse effects reported in either group. Conclusions: Our study suggests that visceral manipulation in addition to standard care is not effective in changing short-term outcomes but may produce clinically worthwhile improvements in pain at 1 year.
1. Background Despite the large number of randomized controlled trials investigating interventions for low back pain (LBP), recommended interventions produce only small effects (Keller et al., 2007). It is possible that outcomes could be improved by combining current interventions or by the addition of new therapies. Visceral manipulation is a manual therapy technique aimed at treating motion abnormalities of the internal organs (Barral, 2005; Vleminckx, 2006). © 2014 European Pain Federation - EFIC®
Although visceral manipulation is not taught in basic physiotherapy training, training in this treatment technique is being sought out by many physiotherapists and health practitioners in post-graduate workshops. However, there is paucity of research into the efficacy of this approach. There is preliminary evidence that visceral manipulation may be effective for treating LBP. A recent clinical series demonstrated improved symptoms following a specific visceral manipulation technique aimed at mobilizing the kidneys in people with nonspecific LBP Eur J Pain •• (2014) ••–••
1
Does visceral manipulation alter low back pain outcomes?
What’s already known about this topic? • Visceral manipulation has been used for over a decade in the treatment of low back pain, but its efficacy remains untested. What does this study add? • This study demonstrates that the addition of visceral manipulation to a standard physiotherapy treatment algorithm improves long-term but not short-term pain.
(Tozzi et al., 2012). Another within-subject, repeated measures study demonstrated immediate lumbar pressure/pain threshold reduction in asymptomatic subjects following visceral manipulation techniques (McSweeney et al., 2012). The rationale for this therapeutic approach is that visceral disorders could potentially trigger or exacerbate LBP symptoms in the presence of impaired movement between internal organs and their supportive tissues. This could manifest as LBP via two possible mechanisms: visceral referred pain and central sensitization. The mechanism by which visceral pain causes referral to somatic structures could occur by neural convergence, whereby sympathetic afferent nerves that convey signals from the viscera converge with somatic nerves in the dorsal horn. Due to the low proportion of visceral to somatic afferents entering the dorsal horn (Cervero, 2000) and because activation of certain visceral receptors does not induce conscious perception (Cervero, 2009), visceral nociceptive input can be misinterpreted as arising from somatic structures. Activation of visceral nociceptors can result in central sensitization. These nociceptors can be stimulated by an altered gut environment or by altered gastrointestinal/urinary motility (Cervero, 2009). Hence, small variations around the visceral sensory receptors can evoke peripheral hypersensitivity (Cervero, 1995). Even a low level of peripheral input can cause increased activity of projection neurons in the spinal cord and can dynamically maintain this central activation (Woolf, 2011). Central sensitization can result in normal sensory stimuli, such as mechanical touch, being experienced as pain (Woolf, 2011). The mechanisms by which visceral manipulation may have an effect on pain are not yet understood. One hypothesis is that by specific manual treatment of the supportive fascia of the internal cavities of the trunk, visceral manipulation modulates visceral nociceptive signalling (Vleminckx, 2006), reducing excessive visceral nociceptive input entering the spinal cord 2 Eur J Pain •• (2014) ••–••
J. Panagopoulos et al.
and allowing hypersensitive central neurons to return to a normal state of excitation (Woolf, 2011). Therefore, it is possible that visceral disorders contribute to the development and continuation of LBP in some patients and that standard therapies which do not directly target the viscera (and surrounding fascia) could neglect a potential pain-producing component. The efficacy of visceral manipulation for LBP has not yet been evaluated in a randomized controlled trial. The aim of this study was to investigate whether the addition of visceral manipulation to standard physiotherapy care for LBP improves pain, disability and functional outcomes.
2. Methods 2.1 Study design The study design was a randomized placebo controlled trial with blinded follow-up. Ethical approval for this study was received from the University of Sydney Human Research Ethics Committee. This study was registered with the Australia and New Zealand Clinical Trials Registry (ACTRN12611000757910) and the study protocol was published (Panagopoulos et al., 2013).
2.2 Study population/recruitment Consecutive patients who presented with LBP to a private physiotherapy clinic in Sydney, Australia were screened for eligibility. Potential participants who met the inclusion criteria were invited to participate in the trial and provided written consent. To be eligible for the trial, participants were required to meet all of the following criteria as assessed by the treating physiotherapist: primary complaint of pain in the area extending from the 12th rib to the buttock crease (this may or may not be accompanied by pain in the leg or other spinal areas); LBP symptoms which have a score of >2/10 on the 0–10 Numerical Pain Rating Scale (NPRS) (Pengel et al., 2004); aged 18–80; no known or suspected serious spinal pathology (e.g., metastatic, inflammatory or infective diseases of the spine, cauda equina syndrome, canal stenosis, spinal fracture); no nerve root compromise evidenced by at least two of the following (1) myotomal weakness, (2) dermatomal or widespread sensory loss, (3) hypoor hyperreflexia of the lower limb reflexes; no spinal surgery within the preceding 6 months; no visceral surgery within the preceding 6 months; no vascular abnormality such as abdominal aortic aneurysms; not currently receiving chiropractic, osteopathic or other physical therapy; not pregnant or suspect being pregnant; not currently in an acute inflammatory phase of known gastrointestinal or urinary diseases (such as cholecystitis, renal calculi, peritonitis, appendicitis); not currently taking medications that significantly alter gut motility; not currently taking medications (such as oral cor-
© 2014 European Pain Federation - EFIC®
J. Panagopoulos et al.
Does visceral manipulation alter low back pain outcomes?
ticosteroids which are known to increase the risk of intestinal perforation); no known gastrointestinal disease that associates with a risk of intestinal perforation (e.g. Crohn’s disease, diverticular disease, peptic ulcer disease); not taking anti-platelet medications such as warfarin.
2.3 Baseline measures and randomization Participants completed baseline data immediately prior to being randomized to a treatment arm of the study (Fig. 1). A total of 98 patients were eligible and 34 were excluded as they did not fit the inclusion criteria (Fig. 1). Of these excluded patients, 18 chose not to participate, 12 had a pain rating which was too low at baseline, two were pregnant, one had had previous lumbar surgery and one patient was over 80 years of age (Fig. 1). At initial assessment, demographic data and the following baseline measures were collected: pain, measured with the 0–10 NPRS (Pengel et al., 2004), where 0 = no pain and 10 = pain as bad as it could be; disability, measured with the 0–24 RolandMorris Disability Scale (Roland and Morris, 1983), where 0 = no disability and 24 = severe disability; function, measured with the 0–10 Patient-Specific Functional Scale (Stratford et al., 1995), in which the patient nominates three important activities which are
limited by their pain and rates their ability to perform them on a Likert scale ranging from 0 (unable to perform the activity) to 10 (able to perform the activity at pre-injury level). The scores are summed and averaged producing a score out of 10; presence of gastrointestinal/urinary/reproductive symptoms was assessed by asking participants if they suffer from bloating, diarrhoea, constipation, period pain, cramping, food sensitivity or reflux. This variable of visceral symptomatology allowed researchers to determine whether patients with obvious visceral symptoms were randomized equally to both treatment arms of the study. A researcher not involved in data collection or analysis developed a randomization schedule using Excel to generate 64 sequentially numbers, which were concealed in sealed, opaque randomization envelopes. These envelopes contained a paper with words ‘VM’ or ‘placebo’. After baseline data had been collected, the treating physiotherapist opened the next randomization envelope and allocated the participant according to the randomization schedule (Fig. 1).
2.4 Clinical assessment Participants were assessed by one of two experienced musculoskeletal physiotherapists. Both physiothera-
98 Eligible LBP patients attend for physiotherapy assessment and treatment
34 patients excluded 18 chose not to participate 12 pain rating too low 2 pregnant 1 previous lumbar surgery
64 participants randomized
1 >80 years of age
32 Placebo group
32 Visceral manipulation group
standard physiotherapy + sham visceral manipulation
standard physiotherapy + real visceral manipulaion
31 followed-up at 2 weeks (97%)
32 followed-up at 2 weeks (100%)
30 followed-up at 6 weeks (94%)
32 followed-up at 6 weeks (100%)
28 followed-up at 52 weeks (88%)
31 followed-up at 52 weeks (97%)
Figure 1 Study flow diagram.
© 2014 European Pain Federation - EFIC®
Eur J Pain •• (2014) ••–••
3
Does visceral manipulation alter low back pain outcomes?
pists had concluded post-graduate intensive practical and coursework training in visceral manipulation and had between 6 and 10 years of clinical experience in the use of visceral manipulation. All participants received a standardized assessment involving active movement testing, passive intervertebral motion testing, palpation, neural tension testing and functional stability testing. Participants also received palpation testing of their visceral system to assess which visceral structures may be involved in the participant’s clinical presentation. This palpation testing involved the gastrointestinal, urinary, respiratory, reproductive and cardiovascular systems as described in detail in Barral (2005). In brief, the physiotherapist placed his hands on the anterior part of the subject’s body and assessed organ motility and visceral ligament mobility. For participants allocated to placebo visceral manipulation, these findings were noted but not used in treatment.
2.5 Treatments Participants in both groups received the same standardized physical examination and standard care. The two treating physiotherapists treated both control and intervention patients. As per LBP guidelines, all participants received current evidence-based advice focusing on remaining active (Koes et al., 2010). Participants received manual therapy, exercise therapy (lumbar muscle re-training and functional exercise prescription) and massage, as required by the treating physiotherapist. To facilitate lumbar muscle re-training, real-time ultrasound guided imaging was used to provide visual feedback and education. All participants were treated one to two times per week for a minimum of one week and a maximum of 12 treatments over 6 weeks. The number of treatment sessions was based on the rate of participants’ symptom reduction. Following the 6-week period, treatment was continued if participants felt their treatment or rehabilitation goals had not been met. Participants continued to receive treatment as per their group allocation and remained blinded. For both groups, initial treatment sessions lasted for approximately 40 min and follow-up sessions lasted approximately 25–30 min. In order to ensure compliance and monitor treatments delivered in the study, therapists kept a record of the number of times the patient attended physiotherapy and recorded details of the treatment including the techniques used. The interventions for each group were as follows: 4 Eur J Pain •• (2014) ••–••
J. Panagopoulos et al.
2.5.1 Standard care plus active visceral manipulation group In addition to standard care (as described above), any fascial restrictions were treated using specific visceral manipulation techniques (Barral, 2005; Tozzi et al., 2012). This took approximately 5–10 min and involved light or deep manual fascial releases and specific organ mobilizations in the thoracic, subdiaphragmatic, abdominal and pelvic areas as appropriate (Barral, 2005). 2.5.2 Standard care plus placebo visceral manipulation group Participants were treated with standard care as defined above. Participants also received a placebo visceral ‘treatment’ which involved approximately 5 min of sham treatment aimed around the abdominal area. This involved light touch and no intention on the part of the physiotherapist to impart any therapeutic technique. This placebo therapy was performed on areas of the abdomen not involved in a mechanical, functional or neurological sense to any visceral issues present. The placebo technique was pilot tested on experienced physiotherapists (who had no prior experience of visceral manipulation) prior to starting the trial. The physiotherapists were unable to distinguish between the placebo and real visceral manipulation.
2.6 Outcome measures Measures of outcome were recorded by an assessor blinded to group allocation. Outcomes were recorded at baseline, 2, 6 and 52 weeks after commencement of treatment. If the participant had ceased treatment, the outcome measures were collected by a blinded assessor over the phone or by email. At 6 weeks, participants were asked a treatment credibility question regarding which additional treatment they thought they received (real or placebo treatment). This was performed to assess if participant blinding was successful. 2.6.1 Primary outcome The primary outcome was pain intensity (0–10 NPRS) at 6 weeks as it was hypothesized to be the time when the intervention may have largest effect. 2.6.2 Secondary outcomes The secondary outcomes were pain intensity (0–10 NRPS) at 2 and 52 weeks, function and disability at 2, 6 and 52 weeks. © 2014 European Pain Federation - EFIC®
J. Panagopoulos et al.
Does visceral manipulation alter low back pain outcomes?
Table 1 Effects of visceral manipulation compared with control.
Visceral manipulation
Control
Adjusted treatment differences (95% confidence interval)
3.06 (2.08) 2.31 (1.99) 1.52 (1.65)
3.74 (2.25) 2.33 (2.22) 3.21 (2.27)
0.55 (−0.65 to 1.75) −0.12 (−1.45 to 1.21) 1.57 (0.32 to 2.82)
0.362 0.858 0.015
5.78 (5.40) 3.00 (2.96) 2.06 (3.56)
6.26 (5.35) 3.10 (3.98) 3.50 (3.61)
−0.86 (−3.33 to 1.58) −1.20 (−4.18 to 1.75) 0.11 (−2.86 to 3.07)
0.479 0.418 0.942
6.10 (2.13) 7.70 (1.81) 8.43 (1.76)
6.15 (1.95) 7.51 (1.86) 7.55 (1.82)
0.81 (−0.33 to 1.94) 0.61 (−0.63 to 1.84) −0.08 (−1.18 to 1.02)
0.158 0.332 0.882
Unadjusted mean outcomes (standard deviation) Outcome Pain (NPRS) 2/52 6/52 52/52 Disability (RMDQ) 2/52 6/52 52/52 Function (PSFS) 2/52 6/52 52/52
p-value
Numerical Pain Rating Scale (NPRS): where 0 = no pain and 10 = pain as bad as it could be. Roland-Morris Disability Questionnaire (RMDQ): where 0 = no disability and 24 = severe disability. Patient-Specific Functional Scale (PSFS): patient nominated three important activities which were limited by their injury and rated these on a Likert scale ranging from 0 (unable to perform the activity) to 10 (able to perform the activity at pre-injury level). The scores were summed and averaged producing a score out of 10.
2.7 Data analysis
3. Results
The sample size of 64 was determined a priori (Panagopoulos et al., 2013). This sample size was calculated to provide 80% power to detect a 1.5 point difference (which we considered to be a worthwhile effect) on the 0–10 NPRS, allowing for an estimated standard deviation of 2 and loss to follow-up of 10%. An ‘unstructured’ covariance matrix was assumed, although the results were very similar using a range of other structures (autoregressive, Toeplitz, HuynhFeldt, compound symmetry). All data were double entered. Data analysis was performed by a statistician who was blinded to group status and following intention-to-treat principles. The number of analyses was limited to reduce the possibility of type I errors. For primary outcomes, a p-value of
