Management of lumbar zygapophysial (facet) joint pain

Table 6 Study characteristics of randomized controlled trials of lumbar radiofrequency neurotomy, facet joint nerve blocks, and intraarticular injections

Study/study characteristic/methodological quality Scoring	No. of patients and selection criteria	Control	Interventions	Outcome measures	Time of measurement	Reported results	Strengths	Weaknesses	Conclusion/comments
Radiofrequency neurotomy
Civelek et al[108], 2012	100 patients with chronic low back pain with failed conservative therapy and strict selection criteria; however, without diagnostic blocks	Facet joint nerve block with local anesthetic and steroids in 50 patients	Conventional radiofrequency neurotomy at 80 °C for 120 s in combination with high dose local anesthetic and steroids, in 50 patients	Visual Numeric Pain Scale, North American Spine Society patient satisfaction questionnaire, Euro-Qol in 5 dimensions and ≥ 50% relief	1, 6 and 12 mo	At one year, 90% of patients in the radiofrequency group and 69% of the patients in the facet joint nerve block group showed significant improvement compared to 92% and 75% at 6-mo follow-up	Randomized relatively large number of patients with 50 in each group	No diagnostic blocks were performed. High dose steroids and local anesthetics were utilized in both groups	Efficacy was shown even without diagnostic blocks, both for facet joint nerve blocks and radiofrequency neurotomy
Randomized, active-control trial
Quality scores:
Cochrane = 9/12
IPM-QRB = 28/48
Cohen et al[109], 2010	151 chronic low back pain, 51 patients with no diagnostic block, 50 patients a single diagnostic block, 50 patients in double diagnostic block	Radiofrequency neurotomy in patients without diagnostic blocks	Conventional radiofrequency neurotomy at 80 °C for 90 s in all patients; however, in 2 groups with either a single block paradigm or a double block paradigm testing for positive results	Greater than 50% pain relief coupled with a positive global perceived effect persisting for 3 mo	3 mo	Denervation success rates in groups 0, 1 and 2 were 33%, 39% and 64%, respectively	Multicenter, randomized controlled trial with or without diagnostic blocks	Authors misinterpreted cost-effectiveness without consideration of many factors reported	Results showed efficacy when double diagnostic blocks were utilized
Randomized, double-blind, control trial
Quality scores:
Cochrane = 8/12
IPM-QRB = 28/48
Nath et al[105], 2008	40 patients with chronic low back pain for at least 2 yr with 80% relief of low back pain after controlled medial branch blocks. The patients were randomized into an active and a control group	Controlled sham lesion in 20 patients in the control group	The 20 patients in the active group received conventional lumbar facet joint radiofrequency neurolysis at 85 °C for 60 s. The 20 patients in the control group received sham treatment without radiofrequency neurolysis of the lumbar facet joints	Numeric Rating Scale, global functional improvement, reduced opioid intake, employment status	6 mo	Significant reduction not only in back and leg pain; functional improvement; opioid reduction; and employment status in the active group compared to the control group	Randomized, double-blind trial after the diagnosis of facet joint pain with triple diagnostic blocks	Short-term follow-up with small number of patients	Efficacy of radiofreqency neurotomy was shown compared to local anesthetic injection and sham lesioning
Randomized, double-blind, sham control trial
Quality scores:
Cochrane = 12/12
IPM-QRB = 42/48
Tekin et al[96], 2007	60 patients with chronic low back pain randomized into 3 groups with 20 patients in each group. Single diagnostic block of facet joint nerves with 0.3 mL of lidocaine 2% with 50% or greater relief	Sham control with local anesthetic injection	Either pulsed radiofrequency (42 °C for 4 min) or conventional radiofrequency neurotomy (80 °C for 90 s) in 20 patients in each group	Visual analog scale and Oswestry Disability Index	3, 6 and 12 mo	Visual analog scale and Oswestry Disability Index scores decreased in all groups from 3 procedural levels. Decrease in pain scores was maintained in the conventional radiofrequency group at 6 mo and one year. However, in pulsed radiofrequency group, the improvement was significant only at 6 mo, but not 1 yr	Randomized, double-blind, controlled trial comparing control, pulsed radiofrequency, and conventional radiofreqency neurotomy. Authors also utilized a parallel needle placement approach	Small sample size with a single block and 50% relief as inclusion criteria. Authors have not described the significant improvement percentages	Efficacy with conventional radiofreqency neurotomy up to one year, whereas efficacy with local anesthetic block with sham control radiofrequency neurotomy and pulsed radiofrequency neurotomy at 6 mo only
Randomized, active and sham, double-blind controlled trial
Quality scores:
Cochrane = 12/12
IPM-QRB = 37/48
van Wijk et al[97], 2005	81 patients with chronic low back pain were evaluated with radiofreqency neurotomy with 41 patients in the control group with at least 50% relief for 30 min with a single block with intraarticular injection of 0.5 mL lidocaine 2%	Sham lesion procedure after local anesthetic injection	40 patients received conventional radiofrequency lesioning at 80 °C for 60 s and 41 patients received sham lesioning	Pain relief, physical activities, analgesic intake, global perceived effect, short-form-36, quality of life measures	3 mo	Global perceived effect improved after radiofrequency facet joint denervation. The visual analog scale in both groups improved. The combined outcome measures showed no difference between radiofreqency facet joint denervation (27.5% vs 29.3% success rate)	Double-blind, sham control, randomized trial	Poor selection with a single diagnostic block of 50% pain reduction even though 17.5% of the patients tested positive. Further, authors described that the needle was positioned parallel; however, the radiographic figures illustrate the needle was being positioned perpendicularly rather than parallel to the nerve	Lack of efficacy with methodological deficiencies and a short-term follow-up
Randomized, double-blind, sham control trial
Quality scores:
Cochrane = 12/12
IPM-QRB = 36/48
Dobrogowski et al[98], 2005	45 consecutive patients with chronic low back pain judged to be positive with controlled diagnostic blocks	Injection of saline in patients after conventional radiofrequency (85 °C for 60 s) neurotomy to evaluate postoperative pain	Conventional radiofrequency neurotomy at 85 °C for 60 s, followed by injection of either methylprednisolone or pentoxifylline	Visual analog scale, minimum of 50% reduction of pain intensity, patient satisfaction score	1, 3, 6 and 12 mo	Greater than 50% of reduction of pain intensity was observed in 66% of the patients 12 mo later. There was no difference in the long-term outcomes	Randomized, active control trial	Very small study evaluating effectiveness of radiofrequency neurotomy and postoperative pain	Radiofrequency neurotomy effective with or without steroid injection after neurolysis
Randomized, active control trial
Quality scores:
Cochrane = 10/12
IPM-QRB = 29/48
van Kleef et al[99], 1999	31 patients with a history of at least one year of chronic low back pain randomly assigned to one of 2 treatment groups. Single diagnostic block with 50% relief	Sham control of radiofrequency after local anesthetic injection in 16 patients	The 15 patients in the conventional radiofrequency treatment group received an 80 °C radiofrequency lesion for 60 s	Visual analog scale, pain scores, global perceived effect, Oswestry Disability Index	3, 6 and 12 mo	After 3, 6 and 12 mo, the number of successes in the lesion and sham groups was 9 of 15 (60%) and 4 of 16 (25%), 7 of 15 (47%) and 3 of 16 (19%), and 7 of 15 (47%) and 2 of 16 (13%) respectively. There was a statistically significant difference	Double-blind, randomized, sham controlled trial	A single block with a small sample with inclusion criteria of 50% pain relief to enter the study. The study has been criticized that electrodes were placed at an angle to the target nerve, instead of parallel (51A)	Efficacy shown in a small sample with a single diagnostic block
Randomized, double-blind, sham control trial
Quality scores:
Cochrane = 12/12
IPM-QRB = 40/48
Moon et al[100], 2013	82 patients were included with low back pain with 41 patients in each group either with a parallel placement of the needle or perpendicular placement of the needle	An active control trial with needle placement with perpendicular approach	41 patients in each group were treated with radiofrequency (80 °C for 90 s) after appropriate diagnosis of facet joint pain with dual diagnostic blocks with 50% relief as the criterion standard. The needle was positioned either utilizing a discal approach or perpendicular or utilizing tunnel vision approach with parallel placement of the needle	NRS, ODI	1 and 6 mo	Patients in both groups showed a statistically significant reduction in NRS and Oswestry disability index scores from baseline to that of the scores at 1 and 6 mo (all P < 0.0001, Bonferroni corrected)	Randomized, double-blind, controlled trial. The major strength is that authors have proven that parallel approach may not be the best as has been described. Diagnosis of facet joint pain by dual blocks	Active controlled trial without placebo group. Short-term follow-up	Positive results in an active controlled trial, in a relatively short-term follow-up of 6 mo, with positioning of the needle either with distal approach (perpendicular placement or tunnel vision) with parallel placement of the needle with some superiority with perpendicular approach. This trial abates any criticism of needle positioning one way or the other and the traditional needle positioning appears to be superior to parallel needle placement
Prospective, randomized, comparative study
Quality scores:
Cochrane = 9/12
IPM-QRB = 38/48
Lakemeier et al[101], 2013	56 patients were randomized into 2 groups with 29 patients receiving intraarticular steroid injections and 27 patients receiving radiofrequency denervation after the diagnosis was made with intraarticular injection of local anesthetic with a single block	Intraarticular injection of local anesthetic and steroid	Radiofrequency neurotomy for 90 s at 80 °C	Roland-Morris questionnaire, VAS, ODI, analgesic intake	6 mo	Pain relief and functional improvement were observed in both groups. There were no significant differences between the 2 groups for pain relief and functional status improvement	Lack of placebo group. Relatively short-term follow-up	Randomized, double-blind trial with single diagnostic block with intraarticular injection	Both groups showed improvement. Effectiveness at 6 mo in both groups with intraarticular injection or radiofrequency neurotomy
Randomized, double-blind, active controlled trial
Quality scores:
Cochrane = 9/12
IPM-QRB = 37/48
Lumbar facet joint nerve blocks
Civelek et al[108], 2012	100 patients with chronic low back with failed conservative therapy and strict selection criteria; however, without diagnostic blocks	Blocks of facet joint nerves with local anesthetic and steroids	Conventional radiofrequency neurotomy at 80 °C for 120 s in combination with high dose local anesthetic and steroids	Visual Numeric Pain Scale, North American Spine Society patient satisfaction questionnaire, Euro-Qol in 5 dimensions and ≥ 50% relief	1, 6 and 12 mo	At the end of one year, 90% of patients in the radiofrequency group and 69% of the patients in the facet joint nerve block group showed significant improvement vs 92% and 75% at 6-mo follow-up	Randomized active-control trial with relatively large number of patients with 50 in each group	No diagnostic blocks were performed. High dose steroids and local anesthetics were provided in both groups	Results showed efficacy even without diagnostic blocks, both for facet joint nerve blocks and radiofrequency neurotomy
Randomized, active-control trial
Quality scores:
Cochrane = 9/12
IPM-QRB = 28/48
Manchikanti et al[102], 2010	120 patients with chronic low back pain of facet joint origin treated with therapeutic lumbar facet joint nerve blocks Double diagnostic blocks with 80% relief	Local anesthetic only	Total of 120 patients with 60 patients in each group with local anesthetic alone or local anesthetic and steroids. Both groups were also divided into 2 categories each with the addition of Sarapin	Numeric Rating Scale, Oswestry Disability Index, employment status, and opioid intake	3, 6, 12, 18 and 24 mo	Significant pain relief was shown in 85% in local anesthetic group and 90% in local anesthetic with steroids group at the end of the 2 yr study period in both groups, with an average of 5-6 total treatments	Randomized trial with relatively large proportion of patients with 2-yr follow-up, with inclusion of patients diagnosed with controlled diagnostic blocks	Lack of placebo group	Effectiveness demonstrated with facet joint nerve blocks with local anesthetic with or without steroids
Randomized, double blind, active control trial
Quality scores:
Cochrane = 11/12
IPM-QRB = 45/48
Lumbar intraarticular injections
Carette et al[103], 1991	Patients with chronic low back pain who reported immediate relief of their pain after injection of local anesthetic into the facet joints. Single diagnostic blocks with 50% relief were randomly assigned to receive injections under fluoroscopic guidance	Intraarticular injection of isotonic saline	Injection of either sodium chloride or methylprednisolone into the facet joints (49 for isotonic saline and 48 for sodium chloride). Only one injection was provided	Visual Analog Scale, McGill Pain Questionnaire, mean sickness impact profile	1, 3 and 6 mo	After 1 mo, 42% of the patients in the methylprednisolone group and 33% in the sodium chloride group reported marked or very marked improvement. At the 6 mo evaluation, 46% in the methylprednisolone group and 15% in the placebo group showed sustained relief. Revised statistics showed 22% improvement in active group and 10% in control group	Well-performed randomized, double-blind controlled trial	Only single block was applied and patients were treated with steroids without local anesthetic with only one treatment and expected 6 mo of relief	The authors concluded that results were negative in an active-control trial with injection of either sodium chloride solution or steroid into the facet joints after diagnosis with a single block
Randomized, double blind, impure placebo or active- control trial
Quality scores:
Cochrane = 11/12
IPM-QRB = 40/48
Fuchs et al[104], 2005	60 patients with chronic low back pain were included with patients randomly assigned into 2 groups. No diagnostic blocks	Active-control study with no control group	Intraarticular injection of hyaluronic acid vs glucocorticoid injection	VAS, Rowland-Morris Questionnaire, ODI, low back outcomes score, short form-36	3 and 6 mo	Patients reported lasting pain relief, better function, and improved quality of life with both treatments	Randomized, active-control, double-blind study	Relatively small sample of patients with 6 mo follow-up without a placebo group, without diagnostic blocks	Undetermined (clinically inapplicable) results with high number of injections during a 6-mo period
Randomized, double-blind, active-control trial
Quality scores:
Cochrane = 8/12
IPM-QRB = 26/48
Ribeiro et al[107], 2013	60 patients with a diagnosis of facet joint syndrome randomized into experimental and control groups	Triamcinolone acetonide intramuscular injection of 6 lumbar paravertebral points	Intraarticular injection of 6 lumbar facet joints with triamcinolone hexacetonide	Pain visual analogue scale, pain visual analogue scale during extension of the spine, Likert scale, improvement percentage scale, Roland-Morris, 36-Item Short Form Health Survey, and accountability of medications taken	1, 4, 12 and 24 wk	Both groups showed improvement with no statistical difference between the groups. Improvement "percentage" analysis at each time point showed significant differences between the groups at week 7 and week 12. Improvement percentage was > 50% at all times in the experimental group with intraarticular steroids; however, significant difference was noted at 24 wk	Randomized, double-blind controlled trial	Diagnostic blocks were not employed, thus, many patients without facet joint pain may have been included in this trial	Overall intraarticular steroids showed positive effectiveness for 24 wk compared to intramuscular steroids provided in a double-blind manner
Randomized, double-blind, active control
Quality scores:
Cochrane = 10/12
IPM-QRB = 32/48
Lakemeier et al[101], 2013	56 patients were randomized into 2 groups receiving intraarticular steroid injections or radiofrequency denervation after the diagnosis was made with intraarticular injection of local anesthetic with a single block	Intraarticular injection of local anesthetic and steroid in 29 patients	Radiofrequency neurotomy for 90 s at 80 °C in 27 patients	Roland-Morris questionnaire, VAS, ODI, analgesic intake	6 mo	Pain relief and functional improvement were observed in both groups. There were no significant differences between the 2 groups for pain relief and functional status improvement	Lack of placebo group. Relatively short-term follow-up	Randomized, double-blind trial with single diagnostic block with intraarticular injection	Both groups showed improvement. Effectiveness of both modalities at 6 mo in both groups
Randomized, double-blind, active controlled trial
Quality scores:
Cochrane = 9/12
IPM-QRB = 37/48
Yun et al[106], 2012	57 patients with facet syndrome were assigned to 2 groups with 32 patients in the fluoroscopy group and 25 patients in the under ultrasonography group without diagnostic blocks	Intraarticular injection of lidocaine and triamcinolone under fluoroscopic guidance	Intraarticular injection of lidocaine and triamcinolone under ultrasonic guidance	VAS, physician’s and patient’s global assessment (PhyGA, PaGA), modified ODI	1 wk, 1 and 3 mo	Each group showed significant improvement from the facet joint injections. However at a week, a mo, and 3 mo after injections, no significant differences were observed between the groups	Randomized trial	Short-term follow-up with no diagnostic blocks, thus increasing the potential for inclusion of patients without facet joint pain. The aim of study mainly was to confirm if ultrasonic imaging was appropriate	The study showed positive results in both groups with intraarticular steroid injections with a short-term follow-up whether performed under ultrasonic guidance or fluoroscopy
Randomized, active controlled trial
Quality scores:
Cochrane = 9/12
IPM-QRB = 26/48

IPM-QRB: Interventional Pain Management Techniques - Quality Appraisal of Reliability and Risk of Bias Assessment; VAS: Visual analog scale; ODI: Oswestry Disability Index.