Table 1 Characteristics about efficacy and safety of included studies

Holanda 2016 [14]

RCT / pilot study

No other intervention

PRF treatment group n = 11;

lidocaine injection group n = 7;

laser irradiation treatment group n = 10

5 min and 1 month

1. Lumbar pain intensity by VAS presented as percentage of relative difference

2. Chronic LBP relief by PRS

PRF group:

VAS relative difference at 5 min: 62.5%

VAS relative difference at 1 month: 20%

Lidocaine injection group: VAS relative difference at 5 min: 100%

VAS relative difference at 1 month: 62.5%

Laser treatment group:

VAS relative difference at 5 min: 100%

VAS relative difference at 1 month: 55%

Positive conclusive

No SAEs occurred

Some patients experienced only mild discomfort during procedure

Not reported

Lee 2018 [15]

RCT / randomized, prospective, and comparative study

Diagnostic block + PRF group received diagnostic block with 1 ml of 2% bupivacaine and 1 ml of 2% triamcinolone.

Diagnostic block + PRF group (n = 30);

PRF group (n = 30)

2 weeks, 1, 3 and 6 months

1. Pain intensity by NRS

2. Functional disabilities by ODI

Diagnostic block + PRF group:

Baseline NRS: 8 (range 5–9);

NRS at 2 weeks: 2 (range 1–7);

NRS at 1 month: 2 (range 1–8);

NRS at 3 months: 3 (range 1–8);

NRS at 6 months: 4 (range 1–8).

PRF alone group:

Baseline NRS: 7.5 range (3–10);

NRS at 2 weeks: 2 (range 1–9);

NRS at 1 month: 2 (range 1–9);

NRS at 3 months: 3 (range 1–9);

NRS at 6 months: 4 (range 1–9).

P values comparison between groups:

NRS at 2 weeks: P = 0.302

NRS at 1 month: P = 0.690

NRS at 3 moths: P = 0.957

NRS at 6 months: P = 0.673

Positive inconclusive

Not reported

Not reported

Not reported

Yang 2010 [16]

RCS / in vivo clinical trial

No other intervention

PaMNI system n = 16;

conventional fluoroscopy n = 13

1 month

1. Pain intensity by VAS

PaMNI group:

Baseline VAS: 5.8 (±2.3);

VAS at 1 month: 4.1 (±2.1).

P = 0.005.

Fluoroscopy group:

Baseline VAS: 6.5 (±2.2);

VAS at 1 month: 5.3 (±2.8).

P = 0.067.

No statistical difference between groups at 1 month (P = 0.238).

Positive conclusive

Not reported

Not reported

Not reported

Hsu 2017 [17]

BA / retrospective study

No other intervention

84

1 week after the treatment and at 3, 6, 9, 12 months and yearly postoperatively (for 3 years in total)

1. Pain intensity by VAS

2. Functional disabilities by ODI

Analysis of VAS scores for pain indicated significant reductions of low back pain during the 3-year follow-up for patients with all 4 types of lumbar lordosis.

Positive conclusive

No SAEs occurred

Cerebral spinal fluid leaking from the cannulas of two patients while the needle was being directed toward the DRG. This leakage ceased immediately after adjusting the location of the needle tip.

Specific adverse events mention, no overall conclusion about safety

Tsou 2010 [18]

BA / not stated

No other intervention

Group A (CLBP without lower-limb pain) n = 49;

Group B (CLBP with lower-limb pain) n = 78

From 1 week up to 3 years post-operatively

1. Pain intensity by VAS

2. Adverse events

Group A, L2 treatment:

≥50% VAS improvement:

at 1 week: 25/49 (51.02%);

at 3 months: 27/49 (55.1%);

at 1 year 20/45 (44.44%).

Group B, L2 treatment:

≥50% VAS improvement:

at 1 week: 34/78 (43.59%);

at 3 months: 37/78 (47.44%);

at 1 year: 34/74 (45.95%).

Positive conclusive

No SAEs occurred

No obvious complications were observed

Positive conclusive

Albayrak 2017 [19]

PCS / retrospective study of prospectively collected

data

In the PRF group participants received prognostic diagnostic block prior to involvement.

PRF group (TENS + exercise + PRF) n = 22;

TENS group (TENS + exercise) n = 17

15 days and 1-month post treatment and following last control examination. The mean follow-up time was 253.8 ± 109 days; for TENS group: 217 ± 114 days and for PRF group: 282.2 ± 97 days

1. Pain intensity by VAS

2. Degree of neuropathic pain reduction by DN4

3. Change in knee flexion by ROM

4. Functional status by WOMAC

5. Patient satisfaction

Success was defined as at least 50% reduction to the VAS (activity, rest, night)

PRF group activity:

baseline VAS: 6.6 (±1.5);

VAS at 15 days: 3 (±1.4);

VAS at 1 month: 3.9 (±2);

VAS at last control: 3.5 (±2.4).

PRF group rest:

baseline VAS: 4.3 (±1.7);

VAS at 15 days: 1.8 (±0.9);

VAS at 1 month: 2.6 (±1.5);

VAS at last control: 2 (±1.6).

PRF group night:

baseline VAS: 3.8 (±2.2);

VAS at 15 days: 1.5 (±1.1);

VAS at 1 month: 2.1 (±1.4);

VAS at last control: 1.7 (±1.4).

TENS group activity:

baseline VAS: 5.9 (±1.9);

VAS at 15 days: 3.8 (±2.2);

VAS at 1 month: 4.3 (±2.2);

VAS at last control: 4.4 (±2.1).

TENS group rest:

baseline VAS: 5 (±2.4);

VAS at 15 days: 2.6 (±2.4);

VAS at 1 month: 3.4 (±2.5);

VAS at last control: 2.8 (±2.1).

TENS group night:

baseline VAS: 4.3 (±2.6);

VAS at 15 days: 2.1 (±2.7);

VAS at 1 month: 2.7 (±2.6);

VAS at last control: 2.6 (±1).

Significant difference achieved in an improvement of at least 50% on the VAS scores at activity following the last control examination between the two groups (P = 0.006), but not on the VAS scores at rest and night (P > 0.05).

Positive conclusive

No SAEs occurred

No complications were observed

Not reported

Cohen 2006 [20]

RCS / retrospective data analysis

No other intervention

PRF DRG group n = 13;

PRF ICN group n = 15;

MM group n = 21

6 weeks, 3 months

1. Pain intensity by VAS

2. Answers to 2 questions evaluating patient satisfaction and functional improvement

Successful was defined as ≥50% pain reduction on VAS and affirmative answer to 2 questions

No separate VAS scores shown in manuscript, success was achieved as follows.:

PRF DRG group:

6 weeks: 61.5%

3 months: 53.8%

PRF ICN group:

6 weeks: 21.4%

3 months: 6.7%

MM group:

6 weeks: 27.3%

3 months: 19.9%

Effect did not reach statistical significance at 6 weeks (P = 0.12).

At 3 months, success rate for PRF DRG group was significantly greater than for those patients treated with PRF ICN (P = 0.01), and approached significance when compared with MM (P = 0.06).

Positive conclusive

Small incidental pneumothorax was found during a routine scan of the lung fields after PRF DRG. This patient was not symptomatic and was treated conservatively with observation.

No other complications occurred

Not reported

Fam 2018 [21]

BA / single arm intervention study

Steroid injections with 1 ml of bupivacaine 0.25% and 1 ml of dexamethasone 4 mg in a total volume of 2 ml immediately after PRF procedure.

n = 100

1 week, 1, 3 and 6 months

1. Pain intensity by VAS

2. Quality of life by QOLS

3. Change in use of pain medication

4. Adverse effects

5. Patient satisfaction

Baseline VAS: 7.48 ± 1.46 (median: 8);

VAS at 1 week: 5.01 ± 2.61 (median: 5) (P = 0.032344);

VAS at 1 month: 3.26 ± 2.37 (median: 3) (P < 0.0001);

VAS at 3 months: 4.44 ± 2.8 (median: 4) (P = 0.00139);

VAS at 6 months: 4.7 ± 2.88 (median: 4) (P = 0.0057).

Positive inconclusive

No SAEs occurred

Pain at the needling site, fever of unknown etiology at the night of intervention, mild to moderate elevation of glucose level in portion of diabetic participants

Positive inconclusive

Kim 2017 [6]

RCS / retrospective cohort study

No other intervention

PRF group n = 20;

continuous epidural group (ropivacaine) n = 22

1, 3 and 6 months

1. Pain intensity by NRS

2. Dose of anticonvulsants and analgesics

PRF group:

baseline NRS: 6.30 ± 0.98

Continuous epidural group:

baseline NRS: 6.73 ± 0.88

NRS values were significantly lower in PRF group from 1  to 3 months and 6 months after the procedure (P = 0.029) than those in continuous epidural group.

Positive conclusive

No SAEs occurred

1 patient complained of pain at the procedure site, and it improved within few days

Not reported

van Zundert 2003 [22]

BA / clinical audit

Diagnostic block prior to involvement.

Participants with > 50% pain relief received PRF.

n = 14

2 months and 6 months after the last patient were included. Mean follow-up was 19.4 months (±8.9 months), maximum 2.5 years.

1.Satisfactory pain relief (GPE: defined as a score of 6 or 7 points on the Likert scale; at least 50% pain relief

2. Duration of the effect

3. Other treatments

4. Change in use of pain medication

Data about pain relief (GPE):

9/14 patients (64%) reported successful pain reduction (6 or 7 points on the GPE Likert scale).

Positive conclusive

No SAEs occurred

No other complications observed

Positive conclusive

Zhang 2011 [23]

CR/CR

Diagnostic blocks with 1.5% lidocaine.

Positive response was considered as 90% pain relief lasting for 30 min.

n = 2

6 months

1. Pain intensity by NRS

Patient 1:

Baseline NRS: 5;

NRS at 6 months: 0.

Patient 2:

Baseline NRS: 4;

NRS at 6 months: 0.

Positive inconclusive

No significant complications occurred

No significant complications occurred

Not reported

Albayrak 2016 [24]

CR / CS

No other intervention

n = 2

1 and 3 days after PRF, 2 and 5 or 10 months (different last follow-up time point for 2 patients)

1. Pain intensity by VAS

2. ROM degree

Patient 1:

Baseline VAS during movement: 80;

VAS at 3 days: 30;

VAS at 2 and 10 months: 20.

Patient 2:

Baseline VAS during movement: 100;

VAS at 3 days: 30;

VAS at 2 months: 20;

VAS at 5 months: 10.

Positive inconclusive

No SAEs occurred

No complications were observed

Not reported

Apiliogullari 2015 [25]

CR / CR

No other intervention

n = 1

1 day after treatment (2 weeks after first PRF the treatment was repeated), 6 months

1. Pain intensity by VAS

Baseline VAS: 100;

VAS at 1 day (PRF on L5): 50;

VAS at 2 weeks (after repeated PRF on L4): 10.

The patient had symptoms relief for over 6 months.

Positive inconclusive

No SAEs occurred

No significant complications occurred

Not reported

Arai 2015 [26]

CS / CR

No other intervention

n = 15

0, 1, 7, 21, 28, 35 and 42 days

1. Pain intensity by NRS at rest and while moving

NRS at rest:

baseline NRS: from 1 to 4 (median 3);

NRS at day 1: median 2;

NRS at day 7: median 1;

NRS at day 21: median 1.

Significant decrease in 3 weeks (P < 0.0001).

NRS while moving:

baseline NRS from 5 to 10 (median 8);

NRS at day 1: median 4;

NRS at day 7: median 4;

NRS at day 21: median 3.

Significant decrease in 3 weeks (P < 0.0001).

Positive conclusive

No SAEs occurred

No other complications occurred

Not reported

Hofmeester 2013 [27]

CR / CR

Diagnostic block with 1 ml of levobupivacaine 0.25%

n = 1

12 months

1. Pain intensity by VAS

Baseline VAS scores: 7–8;

VAS initially after intervention: 4;

VAS at 12 months: 0–1.

Positive conclusive

Not reported

Not reported

Not reported

Lee 2015 [28]

CR / CR

Diagnostic block with 0.3 ml of 0.75% levobupivacaine and 1 mg triamcinolone

n = 1

6 months

1. Pain intensity by VAS

Baseline VAS: 10;

VAS at 6 months: 0.

Positive conclusive

Not reported

Not reported

Not reported

Li 2018 [29]

CR / CR

Diagnostic block with 1 ml of 2% lidocaine

n = 1

1 year

1. Pain intensity by VAS

Baseline VAS: 8;

VAS at 1 year: complete pain relief.

Positive inconclusive

Not reported

Not reported

Not reported