Introduction
Methods
Search strategy
Inclusion criteria
Exclusion criteria
Selection
Data extraction
Risk of bias assessment
Attrition bias | |
Criteria for inclusion explicit | One point |
Listhesis definition appropriately described | One point |
Selection bias | |
Number of men and women given | One point only if gender and age are reported properly |
Age range and mean age reported | |
Can selection-bias be ruled out | One point |
Outcome bias | |
Clearly defined criteria for measuring outcomes | One point |
Patient-reported clinically relevant outcomes | One point |
Preoperative status stated | One point |
Follow-up period, range and mean given | One point |
Valid statistical analysis undertaken | One point |
Clinical evaluation independent of treating physician | One point only if clinical and radiological evaluation were scored independently |
Radiological evaluation independent and blinded to clinical results | |
Randomization bias | |
Allocation bias | One point |
Clear study objective | |
One point | |
Independence of investigators stated | |
One point |
Measures of treatment effect
Results
Search results
Included studies
Study | Study type | Patient characteristics | Treatment |
---|---|---|---|
Prospective studies | |||
Forsth et al. [24] | Type: RCT Selection bias: no Randomization: yes Follow-up: 2 years | 247 patients 155 female/92 male Mean age: 66.7 | The method used for decompression and fusion surgery was determined solely by the surgeon |
Ghogawala et al. [25] | Type: RCT Selection bias: no Randomization: yes Follow-up: 4 years | 66 patients 53 female/13 male Mean age: 66.6 | Group 1: complete laminectomy with partial removal of the medial facet joint Group 2: laminectomy with instrumented fusion |
Herkowitz and Kurz [5] | Type: quasi-RCT Selection bias: no Randomization: yes Follow-up: 4 years | 50 patients 36 female/14 male Mean age: 65 in group 1 and 63.5 in group 2 | Group 1: bilateral laminectomy and facetectomy Group 2: single-level bilateral intertransverse-process arthrodesis |
Matsudaira et al. [26] | Type: prospective cohort study Selection bias: yes Randomization: no Follow-up: 2 years | 37 patients 22 female/15 male Mean age: 68 in group 1 and 67 in group 2 | Group 1: decompression with preservation of the posterior elements Group 2: laminectomy with PLF |
Retrospective studies | |||
Forsth et al. [3] | Type: retrospective cohort study Selection bias: no Randomization: no Follow-up: 2 years | 1306 patients 956 female/350 male Mean age: 69 | Group 1: decompression technique not specified Group 2: decompression and (un)instrumented fusion |
Ghogawala et al. [7] | Type: retrospective cohort study Selection bias: no Randomization: no Follow-up: 1 year | 34 patients 23 female/11 male Mean age: 68.8 | Group 1: decompression technique not specified Group 2: decompression with PLIF |
Kleinstueck et al. [12] | Type: retrospective cohort study Selection bias: no Randomization: no Follow-up: 1 year | 213 patients 155 female/58 male Mean age: 69 | Group 1: laminectomy, flavectomy, lateral recess decompression, partial medial facet resection and very limited laminectomies Group 2: decompression with TLIF, PLIF or PLF |
Park et al. [28] | Type: retrospective cohort study Selection bias: yes Randomization: no Follow-up: 3 years | 45 patients 37 female/8 male Mean age: 64 | Group 1: unilateral laminotomy for bilateral compression (ULBD) Group 2: laminectomy with PLIF |
Plotz and Benini [27] | Type: retrospective cohort study Selection bias: no Randomization: no Follow-up: 3 years | 106 patients 72 female/34 male Mean age: 67 | Group 1: undercutting decompression, laminectomy and decompression Group 2: undercutting decompression or laminectomy combined with translaminar screw fixation or internal pedicle fixator |
Rampersaud et al. [2] | Type: retrospective cohort study Selection bias: yes Randomization: no Follow-up: 2 years | 179 patients 125 female/54 male Mean age: 67.8 in group 1 and 62.47 in group 2 | Group 1: midline-sparing, bilateral decompression from a unilateral approach using a tubular retractor system Group 2: decompression with PLF or PLIF |
Sigmundsson et al. [29] | Type: retrospective cohort study Selection bias: yes Randomization: no Follow-up: 2 years | 836 patients 646 female/190 male Mean age: 71.2 years | Group 1: decompression Group 2: decompression with PLIF |
Risk of bias assessment
Attrition bias (2) | Selection bias (2) | Outcome bias (6) | Randomization (1) | Clear study objective (1) | Independence of investigators (1) | Risk of bias | |
---|---|---|---|---|---|---|---|
Forsth et al. [3] | ** | ** | ***** | – | * | * | Medium |
Forsth et al. [24] | * | ** | ***** | * | * | * | Low |
Ghogawala et al. [7] | * | ** | ***** | – | * | – | Medium |
Ghogawala et al. [25] | ** | ** | ****** | * | * | * | Low |
Herkowitz and Kurz [5] | – | ** | ***** | – | * | * | Medium |
Kleinstueck et al. [12] | * | ** | ****** | – | * | * | Medium |
Matsudaira et al. [26] | * | * | ****** | – | * | – | High |
Park et al. [28] | ** | * | ***** | – | * | * | Medium |
Plotz and Benini [27] | ** | ** | *** | – | * | – | High |
Rampersaud et al. [2] | * | * | ****** | – | * | – | Medium |
Sigmundsson et al. [29] | * | * | ****** | – | * | – | Medium |
Outcome
ODI
Level of evidence
SF-36
Primary outcome | Level of evidence | Study | Results (D = decompression alone, DF = decompression + fusion) | Conclusion |
---|---|---|---|---|
ODIa | Low | Forsth et al. [24] | D: change in ODI (41–21) DF: change in ODI (41–25) | No significant difference (p = 0.11) |
Ghogawala et al. [25] | D (after 2 years): change in ODI (36.3–18.4) D (after 4 years): change in ODI (36.3–21.6) DF (after 2 years): change in ODI (38.8–12.5) DF (after 4 years): change in ODI (38.8–15.1) | No significant difference after 2 years (p = 0.06) and 4 years (p = 0.05) | ||
Forsth et al. [3] | D: change in ODI (45–27) DF: change in ODI (45–27) | No significant difference (p = 0.93) | ||
Ghogawala et al. [7] | D: change in ODI (41.0–27.4; p = 0.003) DF: change in ODI (41.5–14.0; p < 0.001) | Significant difference (p = 0.02) | ||
Park et al. [28] | D: change in ODI (29.8–15.45) DF: change in ODI (24.6–11.0) | No significant difference (p = 0.96) | ||
Sigmundsson et al. [29] | D: predominant leg pain: change in ODI (43.3–25) Predominant back pain: change in ODI (44.9–30.4) DF: predominant leg pain: change in ODI (44.3–24.4) Predominant back pain: change in ODI (46.4–28.8) | Predominant leg pain: no significant difference (p = 0.71) Predominant back pain: no significant difference (p = 0.20) | ||
SF-36b | Very low | Ghogawala et al. [25] | D (after 2 years): change in SF-36 PCS (34.7–44.2) D (after 4 years): change in SF-36 PCS (34.7–42.1) DF (after 2 years): change in SF-36 PCS (31.5–46.7) DF (after 4 years): change in SF-36 PCS (31.5–45.6) | Significant difference after 2 years (p = − 0.046) and 4 years (p = 0.02) |
Ghogawala et al. [7] | D: change in SF-36 PCS (30.9–37.4; p = 0.005) DF: change in SF-36 PCS (29.8–45.7; p < 0.001) | Significant difference (p = 0.003) | ||
Park et al. [28] | D: change in SF-36 PCS (29.2–47.2) and SF-36 MCS (28.0–46.7) DF: change in SF-36 PCS (26.1–46.3) and SF-36 MCS (29.3–44.5) | No significant difference (p = 0.26, p = 0.25) | ||
Rampersaud et al. [2] | D: change in SF-36 PCS (28.89–39.02).and SF-36 MCS (42.91–50.23) DF: change in SF-36 PCS (29.97–41.39 (10.59) and SF-36 MCS (46.78–50.94) | No significant difference (p = 0.39, p = 0.06) | ||
Sigmundsson et al. [29] | D: predominant leg pain: change in SF-36 PCS (37.9–48) and SF-36 MCS (29.2–36.5) Predominant back pain: change in SF-36 PCS (38.2–44.1) and SF-36 MCS (29.0–33.8) DF: predominant leg pain: change in SF-36 PCS (40.7–47.4) and SF-36 MCS (28.2–36.8) Predominant back pain: change in SF-36 PCS (38.6–44.8) and SF-36 MCS (27.5–35.3) | Predominant leg pain: no significant difference (p = 0.16, p = 0.42) Predominant back pain: no significant difference (p = 0.54, p = 0.09) | ||
Back and leg pain | Low/moderate | Forsth et al. [3] | D: VASc back (54–35), VAS leg (63–35) DF: VAS back (61–32) VAS leg (62–32) | No significant difference (back pain: p = 0.12, leg pain: p = 0.17) |
Herkowitz and Kurz [5] | D: change in back pain (2.9–2.5) and leg pain (4.0–1.7) DF: change in back pain (3.3–1.3) and leg pain (4.3–1.0) | Significant difference (p < 0.01). | ||
Kleinstueck et al. [12] | D: reduction back pain: 1.7 (baseline 4.1) and leg pain: 3.1 (baseline 6.5) DF: reduction back pain: 2.9 (baseline 5.3); reduction leg pain: 3.9 (baseline 6.2) | Significant difference reduction in back pain (p = 0.01) No significant difference reduction in leg pain (p = 0.13) | ||
Matsudaira et al. [26] | D: change in back pain (1.4–2.5, p < 0.0001) and leg pain (1.1–2.2, p < 0.0001) DF: change in back pain (1.4–2.4, p = 0.0001) and leg pain (1.0–1.8, p = 0.0001) | No significant difference (p = 0.69, p = 0.52) | ||
Park et al. [28] | D: change in NRSd back (2.8–3.1) and NRS leg (7.8–2.4) DF: change in NRS back (6.6–2.4) and NRS leg (8.0–2.5) | Significant difference NRS back (p = 0.001). No significant difference NRS leg (p = 0.99) | ||
Sigmundsson et al. [29] | D: predominant leg pain: Change in back pain (38.4–30.3) and leg pain (71.3–34.8) Predominant back pain: change in back pain (65.6–38.3) and LP (55.8–35.5) DF: predominant leg pain: Change in back pain (50.8–28.5) and leg pain (73.8–30.8) Predominant back pain: change in back pain (69.4–35.8) and leg pain (56.6–30.8) | Predominant leg pain: significant difference change in back pain (p = 0.0008). No significant difference change in leg pain (p = 0.24) Predominant back pain: no significant differences (p = 0.17, p = 0.38) | ||
Good outcome | Low | Herkowitz and Kurz [5] | D: excellent (8%) good (36%), fair (48%), poor (8%) DF: excellent (44%), good (52%), fair (4%), poor (0%) | Significant difference (p = 0.0001) |
Kleinstueck et al. [12] | D: 70.4% good outcome DF: 86.2% good outcome | Significant difference (p = 0.01) | ||
EQ-5De | No evidence | Forsth et al. [3] | D: EQ-5D (0.36–0.63) DF: EQ-5D (0.33–0.62) | No significant difference (p = 0.34) |
Sigmundsson et al. [29] | D: predominant leg pain: change in EQ-5D (0.34–0.65) Predominant back pain: change in EQ-5D (0.33–0.56) DF: predominant leg pain: change in EQ-5D (0.34–0.66) Predominant back pain: change in EQ-5D (0.30–0.59) | Predominant LP: no significant difference (p = 0.63) Predominant LBP: no significant difference (p = 0.41) | ||
COMIf | No evidence | Kleinstueck et al. [12] | D: reduction in COMI: 3.1 (baseline 7.0) DF: reduction in COMI: 4.2 (baseline 7.6) | Significant difference (p = 0.009) |