Skip to main content
. 2016 Jul 28;2016(7):CD004598. doi: 10.1002/14651858.CD004598.pub4

Price 1998.

Study characteristics
Methods Quasi‐randomised controlled trial (cross‐over)
Participants N = 7 (3 lower extremity, 4 upper extremity pain)
Mean age: 42 years (SD 11; range: 32‐52)
Sex: 3 male
CRPS‐I or ‐II of upper or lower extremities (excluded if CRPS in multiple areas)
Diagnostic criteria: IASP (Merskey 1994)
Mean duration of symptoms: 3 years (SD 2 years; range 18 months to 7 years)
Inciting event: trauma (n = 6), surgical (n = 1)
Medico‐legal factors: not reported
Previous treatment: not reported
Concomitant treatment: all participants continued concomitant physical therapy and medications
Interventions Active condition:
SGB with lidocaine (15 ml of lidocaine 1%)
Lumbar sympathetic blockade with 15 ml 1% lidocaine (test solution) followed by 10 ml bupivacaine 0.125%
Evaluation of technical adequacy of block? yes ‐ evaluated Horner's syndrome and surface skin temperature for stellate ganglion blocks. Nothing reported for lumbar blocks.
Control condition:
Stellate ganglion: 15 ml saline
Lumbar: 15 ml saline + 10 ml saline
The blocks were separated by a period of 7‐10 days
Number of blocks: 1 for each condition
Outcomes Pain intensity and pain unpleasantness (0‐10 VAS)
Pain outcomes measured every 15 min for 1.5 h prior to injection and every 15 min for 1 h following injection. Pain outcomes then rated 4 times a day (morning, midday, afternoon, evening) for 7 days postinjection
Time to peak analgesia measured as the VAS unit difference between pre‐injection baseline pain rating and the lowest VAS rating in the first hour
Duration of pain relief measured as the time it took for pain intensity to return to 50% of the difference between baseline and peak analgesic effect
Country of origin USA
Study aim To evaluate the diagnostic and therapeutic value of local anaesthetic sympathetic blocks
Notes Author confirmed the quasi‐random allocation in correspondence; adverse events not reported
Risk of bias
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) Unclear risk Quote: "Four patients received S first with LA block second, and the order was reversed for the remaining 3 patients"
Comment: quasi‐random process used, not truly random. Due to cross‐over study design and successful blinding, we feel this presents an unclear risk of bias.
Blinding of participants and personnel (performance bias)
All outcomes Low risk Quote: "Both the patient and the physician administering the sympathetic ganglia injections were blind with regard to the contents of the injecting syringe (S or LA) and with regard to whether skin surface temperature changes or Horner's syndrome occurred"; "The syringe was filled. . . by a third person who maintained the code for the contents of the syringes and the double‐blind nature of the study"; "None of the 7 patients reported subjective differences between effects of S and LA blocks within the first hour after block. However, 2 patients correctly determined that they had received S injection because of the shorter duration of relief received."
Comment: blinding completed and blinding success was formally assessed
Blinding of outcome assessment (detection bias)
All outcomes Low risk Outcomes were self rated and participants were blinded to treatment group
Incomplete outcome data (attrition bias)
All outcomes Low risk 2/7 participants missed all pain unpleasantness data but pain intensity data complete
Selective reporting (reporting bias) High risk Group results for pain unpleasantness scores not reported
Adequate sample size? High risk N = 7
Adequate duration of follow‐up? High risk 7 days follow‐up
Free of other bias? Low risk Quote: "Medication use and physical therapy were as similar as possible for the time periods following both saline and lidocaine blocks, and medications were not, as a rule, significantly adjusted during the study period."
Comment: also, cross‐over study design ensured similarity between groups for important outcomes and outcome assessment at same time periods
Procedures separated by 7‐ ‐ 10 days. Figures illustrate pain returned to baseline levels prior to next block.