Table 1. Summary of the literature reviewed in this article investigating the treatment of phantom limb pain.

Authors, year, and sample size	Site (and percentage) of amputation	Reason(s) for amputation and percentages	Mean time (range) since amputation	Treatment groups	Treatment duration	Main outcome measure(s)	Assessments	Findings	Risk of bias
Dumanian et al. 9 (2019) (n = 28)	UE = 13% LE = 87%	Trauma = 90% Infection = 10%	(Less than 1 year to more than 10 years)	1. Targeted muscle reinnervation 2. Standard neuroma surgery	N/A	Change in NRS worst pain from baseline to 12 months post-operatively for PLP and residual limb pain	Baseline 3 months 6 months 9 months 12 months	No significant between- group differences in worst PLP or residual limb pain 1 year post- surgery	High risk
Malavera et al. 16 (2016) (n = 54)	LE = 100%	Trauma = 100%	7.8 years	1. Active rTMS 2. Sham rTMS	20 stimuli of 6 seconds each (54- second intervals), 5 days per week for 2 weeks	PLP intensity measured via a visual analogue scale (VAS)	Baseline 15 days post- rTMS 30 days post- rTMS	No significant between- group differences in PLP scores at either follow-up time	Low risk
Brunelli et al. 19 (2015) (n = 40)	AK = 73% BK = 27%	Dysvascular = 70% Other = 30%	458 days	1. SAIPAN protocol 2. Standard treatment	~1 hour two times per week for 4 weeks	1. PLP intensity, rate, duration, and bother measured via Prosthesis Evaluation Questionnaire (PEQ) and Brief Pain Inventory (BPI) 2. Phantom limb symptom (PLS) intensity, rate, and bother measured via PEQ	Baseline 1 month 2 months	Significant group differences in PLP rate, duration, and bother from PEQ at 2-month follow-up only. No significant differences with BPI. Significant group differences in PLS rate, intensity, and bother at 2-month follow-up only.	High risk
Finn et al. 20 (2017) (n = 15)	AE = 40% BE = 60%	Trauma = 100%	4.5 months (0.55–24 months)	1. MT 2. Control (covered mirror) 3. Control (mental visualization)	15 minutes 5 days per week for 4 weeks	PLP intensity measured via VAS	Baseline 4 weeks	No between-group comparisons reported for main outcome measure	High risk
Anaforoğlu Külünkoğlu et al. 21 (2019) (n = 40)	BK = 100%	Trauma = 100%	13.25 months (3–53 months)	1. MT 2. Phantom exercise (PE)	15 minutes of MT daily at home for 4 weeks PE group performed exercises daily with 15 repetitions	1. PLP intensity measured via VAS 2. Quality of life evaluated via SF-36 3. Psychological status measured using BDI	Baseline 4 weeks 3 months 6 months	Significant group difference in VAS PLP severity, BDI scores, and PF, SF, MH, and V subscales of the SF-36 in favour of the MT group at all follow-up times	High risk
Ol et al. 22 (2018) (n = 45)	BK = 100%	Trauma = 100%	(15–32 years)	1. MT 2. Tactile 3. Combination (mirror and tactile therapy)	5 minutes every morning and night for 4 weeks	PLP intensity measured via VAS	Baseline 5 weeks 3 months after end of treatment	No significant between- group differences in PLP scores at either follow-up time	High risk
Ramadugu et al. 23 (2017) (n = 60)	AE = 8% BE = 8% AK = 34% BK = 50%	NR	NR	1. MT 2. Control (covered mirror)	15 minutes every day for 4 weeks	PLP intensity measured via VAS and the short form of the McGill Pain Questionnaire	Baseline 4 weeks 8 weeks 12 weeks 16 weeks (20 weeks, control only)	No between-group comparisons reported for main outcome measures	High risk
Tilak et al. 24 (2016) (n = 26)	UE = 26% LE = 74%	NR	45 days	1. MT 2. Contralateral transcutaneous electrical nerve stimulation	20 minutes every day for 4 days	PLP intensity measured via VAS and Universal Pain Score (UPS)	Baseline 4 days	No significant between- group differences in PLP intensity at the end of treatment	High risk
Rothgangel et al. 25 (2018) (n = 75)	AK = 61% K = 7% BK = 32%	Trauma = 32% Dysvascular = 40% Tumor = 14% Other = 14%	~35 months	1. MT followed by teletreatment using augmented reality 2. MT followed by self-delivered MT 3. Sensomotor exercises to the intact limb followed by self-delivered exercises	At least 10 30-minute sessions across 4 weeks followed by 6 weeks of self- delivered treatment	PLP frequency, duration, and intensity (measured via NRS)	Baseline 4 weeks 10 weeks 6 months	No significant between- group differences at 4- or 10-week follow-up. At 6-month follow-up, there were significant between- group differences on PLP duration only, in favor of MT followed by self-delivered MT.	High risk
Rostaminejad et al. 26 (2017) (n = 60)	AK = 40% BK = 60%	Diabetes = 45% Trauma = 50% Cancer = 5%	(2–38 months)	1. EMDR 2. Routine care	12 one-hour sessions over 1 month	PLP intensity measured via the subjective units of distress scale and pain rating scale	Baseline 1 month 24 months	No between-group comparisons reported for main outcome measures	High risk

Results from interventions using targeted muscle reinnervation, repetitive transcranial magnetic stimulation (rTMS), mirror therapy (MT), augmented reality, and eye movement desensitization and repossessing (EMDR) therapy are summarized above. For each study, the site, reason, and mean time since amputation are shown. Treatment groups and duration as well as main outcome measures, times of assessment, and major findings are included. The final column describes the article’s global risk-of-bias rating as assessed by the Cochrane Collaboration’s tool ⁸. UE and LE refer to upper and lower extremity amputations, respectively. AK, BK, AE, and BE refer to amputations performed above (A) and below (B) the knee (K) and elbow (E) joint. NR denotes information that was not reported by the author. NRS refers to the numeric 0–10 rating scale. BDI, Beck Depression Inventory; MH, mental health; N/A, not applicable; PF, physical functioning; PLP, phantom limb pain; SAIPAN, Santa Lucia Alleviation Intervention for Phantom in Amputees’ Neurorehabilitation; SF, social functioning; SF-36, 36-Item Short Form Survey; V, vitality.