Schwenk 2015.

Study characteristics
Methods	Study design: randomised controlled trial Study grouping: parallel group
Participants	Baseline characteristics Usual care Age (yr): 71.82 (SD 8.85) Number of women (%): 6 (54.4) Exercise Age (yr): 68.73 (SD 8.72) Number of women (%): 7 (63.6) Inclusion criteria: aged ≥ 55 years; able to provide written consent; diagnosis of current or previous cancer; ability to walk to 10 m without assistive device; presence of chemotherapy‐induced peripheral neuropathy as confirmed by symptoms (numbness, tingling or pain) and signs (reduced vibration perception threshold > 25 V) Excluded criteria: diabetes, foot ulcers or infection; neurological issues (e.g. Parkinson's disease, stroke or multiple sclerosis); severe visual impairment Group differences: none
Interventions	Usual care Asked not to change their physical activity during the study Exercise 2 × 45 minutes × 4 weeks of balance training
Outcomes	Medio‐lateral sway – eyes open Outcome type: continuous Reporting: full Scale: distance Unit of measure: centimetres Direction: lower was better Data value: change from baseline
Identification	Sponsorship source: Flinn Foundation, and National Institute on Aging Country: USA Setting: University of Arizona Cancer Center Comments: none Author's name: Michael Schwenk Institution: Interdisciplinary Consortium on Advanced Motion Performance, University of Arizona E‐mail: schwenk.michael@gmail.com Address: 1501 N Campbell Ave, AHSC 4303D, Tucson, AZ, USA
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Participants were randomly assigned to groups after completion of baseline measurements by a person unrelated to the study using the urn design
Allocation concealment (selection bias)	Low risk	Randomisation was conducted by a person not involved in the study after baseline measurements were completed
Blinding of participants and personnel (performance bias) All outcomes	High risk	Participants were not blinded to their allocation
Blinding of outcome assessment (detection bias)	Low risk	Investigators were unaware of group allocation
Incomplete outcome data (attrition bias) All outcomes	Low risk	Missing outcome data described
Selective outcome reporting (reporting bias)	Low risk	Outcome data checked against protocol registered with ClinicalTrials.gov
Size of study	High risk	Fewer than 50 participants per treatment arm
Other sources of bias	Low risk	Study appeared free of other sources of bias