Jayaraj 2015.

Methods	Study design: randomised controlled trial Method of randomisation: sealed envelopes, with separate blocks for each of the 2 participating institutions Concealment of allocation: yes Blinding: single‐blind. The investigator scoring the presence and severity of PTS was blinded to treatment allocation and results of previous measurements. Follow‐up: 2 years Losses to follow‐up: not accurately described. Seventeen in the treatment group were lost to follow‐up or died; 20 in the non‐intervention group were lost to follow‐up or died.
Participants	Country: United States Number of participants: 69 Age: mean age 48 years Sex: 34 (49%) female Inclusion criteria: acute proximal DVT of the lower extremity documented by duplex ultrasonography. In patients with bilateral acute proximal vein thrombosis, the limb with more extensive disease was designated as the study limb. Exclusion criteria: inability to provide consent, return for follow‐up, or be followed in an anticoagulation clinic; previous DVT, DVT isolated to the calf veins, contraindications to standard anticoagulation, arterial insufficiency (ABI < 0.6), pre‐existing CVI (CEAP class C4 to C6), and current use of compression stockings
Interventions	Intervention group (n = 36): graduated compression stockings (below knee, 30 to 40 mmHg) Control group (n = 33): no stockings Time from diagnosis to randomisation (start of compression stockings): within 48 hours
Outcomes	Primary outcome: assessment of the impact of graduated compression stockings on prevention of PTS Secondary outcomes: comparison of different instruments used to assess PTS including VCSS and VPS instruments
Notes	Complete published version of Jayaraj 2011, including results
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	No description of random sequence generation Quote:"Consenting patients were randomised at the time of enrolment by means of sealed envelopes, with separate blocks for each of the two participating institutions affiliated with the University of Washington."
Allocation concealment (selection bias)	Low risk	Quote:"Consenting patients were randomised at the time of enrolment by means of sealed envelopes, with separate blocks for each of the two participating institutions affiliated with the University of Washington."
Blinding (performance bias and detection bias) Objective outcome(s)	Low risk	Quote: "Patients were instructed not to wear their stockings at their initial or follow‐up clinic visits and not to reveal their treatment allocation at such visits. Each assessment was performed by an investigator (different from the investigator who initially saw patient at the time of randomisation) who was aware of the side of the index DVT but was blinded to treatment allocation or results of previous measurements."
Blinding (performance bias and detection bias) Subjective outcome(s)	High risk	The outcome PTS was based on different scoring systems containing both subjective and objective criteria. Subjective criteria, most of which were patient‐reported items, are at high risk of bias because participants were not blinded. For objective criteria, risk of bias is rather low because of the above mentioned investigator blinding. However, the stockings could leave typical imprints on the skin, which could lead to breaking of blinding.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	High amount of attrition in both groups, with the control group having greatest loss to follow‐up. Kaplan‐Meier analysis was conducted to deal with censored cases.
Selective reporting (reporting bias)	Low risk	All prespecified outcomes were reported.
Other bias	Low risk	Study seems to be free of other risk of bias.