Background
Methods
Types of studies
Types of participants
Types of intervention
Types of outcome measures
Data sources and search strategy
Selection of studies for inclusion and assessment of method quality
Data extraction
Data synthesis
Calculating the efficacy of targeting treatment
Results
Search yield
RCT | Brennan 2006[18](Subacute NSLBP) | Childs 2004[19](Acute NSLBP) | Hancock 2008[26](Acute NSLBP) | Long 2004[21](Chronic NSLBP) |
---|---|---|---|---|
Inclusion criteria
| Age 18-65 years. Low back pain (LBP) of less than 90 days with or without referral into the lower extremity, and an Oswestry disability score ≥ 25%. | Age 18-60 years. Primary symptom of LBP, with or without referral into the lower extremity, Oswestry disability score of at least 30%. | LBP of < 6 weeks duration, causing moderate pain and moderate disability (measured by adaptations of items 7 and 8 of the SF-36) | Age 18-65 years. LBP with or without leg symptoms and with or without a neurological sign. Demonstrating a directional preference. |
Exclusion criteria
| A visible lateral shift or acute kyphotic deformity, signs of nerve root compression, red flags indicating a serious pathology, an inability to reproduce any symptoms with lumbar spine active range of motion (AROM) or palpation, pregnancy, prior surgery to the lumbar and/or sacral region. | Patients with red flags for a serious spinal condition, signs consistent with nerve root compression, pregnancy, prior surgery to the lumbar spine or buttock. | Current episode not preceded by a pain-free period of at least one month in which no care was provided, known or suspected serious spinal pathology, nerve root compromise, currently receiving non-steroidal anti-inflammatory drugs or spinal manipulative therapy, surgery within the preceding 6 months, contraindication to paracetamol, diclofenac or spinal manipulative therapy. | Cauda equina syndrome. Two or more neurological signs. Spinal fractures. Post-surgical. Off work for one year or more due to LBP. Medical causes (for example, severe osteoporosis, inflammatory or infectious conditions). Uncontrolled medical conditions (for example, diabetes, angina, hypertension). Pregnancy. Inability to read English. Patients with prior knowledge of, or specific physician referral for, the McKenzie method. No directional preference. |
Clinical prediction rule
| Delitto Treatment Based Classification | Flynn manipulation prediction rule | Flynn manipulation prediction rule | McKenzie directional preference-based exercise |
Targeted treatment
| Mobilization (low amplitude), manipulation (thrust), exercise (AROM, McKenzie or strengthening and stabilization) n = 50 | Manipulation (thrust), Exercise (ROM) n = 70 | Mobilization (mostly low velocity spinal mobilization, but 5% received manipulation) n = 114 | Exercise (McKenzie directional preference exercises) n = 70 |
Non-targeted treatment
| Mobilization (low amplitude), manipulation (thrust), exercise (AROM, McKenzie or strengthening and stabilization) n = 73 | Exercise (stabilization, low-stress aerobic, strengthening) n = 61 | Sham mobilization (detuned ultrasound) n = 121 | Exercise (exercises opposite to directional preference or non-directional exercises) n = 131 |
Outcomes
| Oswestry Disability Index | Oswestry Disability Index | Roland Morris Disability Questionnaire, Pain Numerical Rating Scale | Roland Morris Disability Questionnaire, Pain Visual Analogue Scale |
Study | Reason for exclusion (studies may have also met other exclusion criteria) |
---|---|
Browder DA, et al. (2007)[37] | Not a two-group plus subgroup covariate RCT or multi-arm subgroup system RCT |
Cairns MC, et al. (2006)[38] | Not a trial of targeted versus non-targeted manual therapy or exercise |
Celestini M, et al. (2005)[39] | Not a trial of targeted versus non-targeted manual therapy or exercise |
Cherkin DC, et al. (1998)[40] | Not a trial of targeted versus non-targeted manual therapy and/or exercise |
Childs JD, et al. (2003)[41] | Not an RCT |
Chiradejnant A, et al. (2002)[42] | Not a two-group plus subgroup covariate RCT or multi-arm subgroup-system RCT |
Chiradejnant A and Kanlayanaphotporn R (2005)[43] | Conference abstract only |
Chiradejnant A, et al. (2003)[44] | More than 15% with neurological signs |
Clare HA, et al. (2007)[45] | No relevant outcome measures |
Descarreaux M, et al. (2002)[46] | Not a two-group plus subgroup covariate RCT or multi-arm subgroup system RCT |
Elnaggar IM, et al. (1991)[47] | Not a trial of targeted versus non-targeted manual therapy and/or exercise |
Erhard, RE, et al. (1994)[48] | Targeted versus targeted treatment |
Fritz JM, et al. (2003)[20] | Not a two-group plus subgroup covariate RCT or multi-arm subgroup system RCT |
Fritz JM, Whitman JM, Childs JD. (2005)[49] | Post-hoc analysis (hypothesis-setting) |
Fritz JM, et al. (2007)[50] | More than 15% with neurological signs |
Geisser ME, et al. (2005)[51] | Not a two-group plus subgroup covariate RCT or multi-arm subgroup system RCT |
Gillan MG, et al. (1998)[52] | Not NSLBP |
Goodsell M, et al. (2000)[53] | Targeted versus no treatment |
Greenman PE (1996)[54] | Not an RCT |
Hough E, et al. (2007)[55] | Not a trial of targeted versus non-targeted manual therapy and/or exercise |
Konstantinou K, et al. (2007)[56] | Cross-over trial, effect size diluted |
Mayer JM, et al. (2005)[57] | Not a trial of targeted versus non-targeted manual therapy and/or exercise |
Miller ER, et al. (2005)[58] | Baseline-scores are not similar between groups (T-test) |
Monticone M, et al. (2004)[59] | Effects of manual therapy, traction or exercise not reported independently of other treatments |
Mujic, SE, et al. (2004)[60] | Not a trial of targeted versus non-targeted manual therapy and/or exercise |
Newton WP (1995)[61] | Not an RCT |
North American Spine Society Board of Directors (2003)[62] | Not an RCT |
O'Brien N, et al. (2006)[63] | Not a trial of targeted versus non-targeted manual therapy and/or exercise |
O'Sullivan PB, et al. (1997)[64] | Not a two-group plus subgroup covariate RCT or multi-arm subgroup system RCT |
Petersen T, et al. (2002)[65] | Not a two-group plus subgroup covariate RCT or multi-arm subgroup system RCT |
Petersen T, et al. (2007)[66] | Not a two-group plus subgroup covariate RCT or multi-arm subgroup system RCT |
Riipinen M, et al. (2005)[67] | Hypothesis-generating study, not hypothesis-testing study |
Rossignol M, et al. (2000)[68] | No relevant intervention |
Schenk RJ, et al. (2003)[69] | More than 15% with neurological signs |
Skikiæ EM, et al. (2004)[70] | Not a trial of targeted versus non-targeted manual therapy and/or exercise |
Spratt, KF, et al. (1993)[71] | Effects of manual therapy, traction or exercise not reported independently of other treatments |
Stankovic R, Johnell O (1990)[72] | Not a trial of targeted versus non-targeted manual therapy and/or exercise |
Sweetman BJ, et al. (1993)[73] | Not a trial of targeted versus non-targeted manual therapy or exercise |
Wright A, et al. (2005)[74] | Not a trial of targeted versus non-targeted manual therapy and/or exercise |
Quality assessment
Method criteria | |||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Randomization | Concealed allocation | Baseline equivalence | Patient blinding | Clinical blinding | Outcome blinding | Co-interventions | Compliance | Dropouts | Outcome timing | Intention to treat | Score | High quality* | |
Brennan 2006[18] | √ | √ | X | √ | √ | √ |
?
| X | X | √ | √ | 7 | Yes |
Childs 2004[19] | √ | √ | √ | √ | √ | √ | X | ? | √ | √ | √ | 9 | Yes |
Hancock 2008[26] | √ | √ | √ | √ | X | √ | √ | √ | √ | √ | √ | 10 | Yes |
Long 2004[21] | √ | √ | √ | X | X | √ | √ | X | √ | ? | √ | 7 | Yes |
Effects of targeting treatment
Brennan 2006[18] |
Delitto Treatment-Based Classification system
Baseline examination data were used to classify participants into one of three classification subgroups: Specific exercise: centralize with two or more movements in the same direction (that is, Flexion or extension) or centralize with a movement in one direction and peripheralize with an opposite movement. Manipulation: onset of symptoms <16 days and no symptoms distal to the knee. Stabilization: at least three of the following: Average straight leg raise range of movement >91 degrees, positive Prone Instability Test, positive aberrant lumbar spine movement, age <40. (Traction was a potential fourth group, but was not included in this study). |
Childs 2004[19] |
Flynn manipulation prediction rule
Patients were classified as positive (likely to respond to manipulation) if they met at least four of these five criteria: Duration of current episode of symptoms less than 16 days, location of symptoms not extending distal to the knee, score on the Fear Avoidance Beliefs Questionnaire (work subscale) less than 19 points, at least one lumbar spine segment judged to be hypomobile, at least one hip with more than 35° of internal rotation range of motion. |
Hancock 2008[26] |
Flynn manipulation prediction rule
As above (Flynn 2003)[35]. |
Long 2004[21] |
McKenzie directional preference-based exercise
Patients were classified as having an extension, flexion or lateral directional preference. |
Outcomes | Mean duration of pain | Three way test of interaction statistically significant* | Mean effect of targeting treatment (95% confidence interval) (0-100 scale) (positive result favours targeted treatment) Bolded scores are statistically significant |
---|---|---|---|
McKenzie directional preference-based exercises | |||
Short term activity limitation
| |||
Directional preference matched exercises versus non-directional preference exercises (Long et al. 2004)[21] | Chronic | NA†
|
16.95 [8.74, 25.16]
P = 0.000, n = 201 |
Short term pain
| |||
Directional preference matched exercises versus non-directional preference exercises (Long et al. 2004)[21] | Chronic | NA |
19.80 [14.34, 25.26]
P = 0.000, n = 201 |
Delitto Treatment-Based Classification
| |||
Short term activity limitation
| |||
Treatment matched to classification vs. treatment unmatched to classification (Brennan 2006)[18] | Sub-acute | Yes | 5.60 [-0.49, 11.69]
P = 0.070, n = 123 |
Long term activity limitation
| |||
Treatment matched to classification vs. treatment unmatched to classification (Brennan 2006) [18] | Sub-acute | Yes | 3.10 [-3.13, 9.33]
P = 0.330, n = 123 |
Flynn manipulation rule
| |||
Short term activity limitation
| |||
Manipulation (fitted prediction rule) versus manipulation (didn't fit rule) (Childs et al. 2004)[19] | Acute | Yes | 8.68 [-1.63, 19.0]
P = 0.10, n = 131 |
SMT (fitted prediction rule) versus SMT (did not fit prediction rule) (Hancock et al. 2008)[26] | Acute | No | -5.50 [-16.09, 5.09] (rule-negative group had better outcome)
P = 0.310, n = 235 |
Intermediate term activity limitation
| |||
Manipulation (fitted prediction rule) versus manipulation (did not fit rule) (Childs et al. 2004)[19] | Acute | Yes | 3.51 [-6.26, 13.28]
P = 0.480, n = 131 |
SMT (fitted prediction rule) versus SMT (did not fit prediction rule) (Hancock et al. 2008)[26] | Acute | No |
-10.30 [-20.80, 0.20]
(rule-negative group had better outcome)
P = 0.050, n = 235 |
Short term pain
| |||
SMT (fitted prediction rule) versus SMT (did not fit prediction rule) (Hancock et al. 2008)[26] | Acute | No | 5.60 [-5.48, 16.68]
P = 0.320, n = 235 |
Intermediate term pain
| |||
SMT (fitted prediction rule) versus SMT (did not fit prediction rule) (Hancock et al. 2008)[26] | Acute | No | 0.40 [-9.84, 10.64]
P = 0.940, n = 235 |