Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Soft collars for whiplash - Bandolier review

Systematic review

Rear impact road traffic accidents are no joke, as Bandolier discovered recently in a misunderstanding with a 40-tonne truck on the M25 around London. One feature is that some people have a soft tissue sprain to the neck, often called a whiplash injury. These are often treated with a soft collar to partially immobilise the neck. A Bandolier reader asked about the evidence for their effectiveness.

Systematic review

Bandolier sought randomised trials comparing soft collars with any other therapy, or no therapy, published in the last 10 years or so. The choice of last 10 years was to minimise any change in soft collar design that may have taken place. It was unlikely that such trials would be blind, and all would likely have low quality scores, so quasi-randomised trials were also accepted.

Any useful outcome was acceptable, but those of greatest interest were pain, time off work, or outcome relating to recovery, at any time after the injury.


Five trials were found (Table 1). Two were quasi-randomised, and two had six-week outcomes. Others had longer outcomes, between six months and three years. Six month or one year outcomes were chosen. Most of the studies used a whole series of outcomes, mostly including pain, but also time off work and other measures. Each had some dichotomous outcome equating to a treatment success, such as no or lower level pain, or no symptoms, or global improvement.

Table 1: Randomised or quasi-randomised trials comparing soft collars with no-treatment or other controls for neck strain after motor vehicle accident

Reference Design Patients Outcomes Results
Gennis et al. Academic Emergency Medicine 1996 3: 568-573 Quasi-randomised (record number).
Soft collar to be worn for as much as possible for first two weeks
No treatment
Both groups advised to rest and use NSAIDs
All patients with neck pain within 24 hours of motor vehicle collision
Six-week report of pain by telephone call
250 patients enrolled, 196 providing data
No pain or better in:
89/104 soft collar
73/92 in control
Borchgrevink et al. Spine 1998 23: 25-31 Random assignment to:
soft collar and 14 days sick leave
act as usual without sick leave
Patients with neck sprain caused by car accident, aged 18-70 years Range of outcomes at six months, including pain in a variety of circumstances, ear symptoms, headache. 201 recruited, with 23 not reporting at six months Global improvement of fewer symptoms
60/96 soft collar
54/82 as usual
Significantly better for act as usual for subjective symptoms, pain, and headache
Rosenfeld et al. Spine 28 22: 2491-2498 Randomised to:
soft collar and written advice
exercise programme based on early and repeated movement
Consecutive patients exposed to whiplash trauma in motor vehicle accidents Range of outcomes at six months and three years.
44 reported who received early intervention in first 96 hours
Low pain at 6 months:
7/23 soft collar
11/21 active
Low pain at 3 years:
9/21 soft collar
7/18 active
Active exercise patients had considerably less sick leave over three years
Schnabel et al. Emergency Medicine Journal 2004 21: 306-310 Randomised to:
soft collar for one week, advised to wear continuously
instruction from physiotherapist on exercises and mobilisation
Patients with pain, stiffness or numbness in spine head or limbs within 48 hours of motor vehicle collission, at least 18 years old Range of outcomes at six weeks.
200 patients randomised; withdrawals 36% of collar group and 15% physiotherapy group
No symptoms at six weeks (per protocol):
27/62 collar
57/88 exercise
No symptoms at six weeks (ITTl):
27/97 collar
57/103 exercise
Exercise group had significantly less pain in neck,and shoulder, and fewer had headaches
Crawford et al. Injury 2004 35: 891-895 Initial treatment with soft collar and NSAIDs, then at clinic randomised (record number) to:
sot collar for 3 weeks then mobilised
early mobilisation without collar
All patients in road traffic accident with neck pain within 48 hours of injury, over age 18 years Activities of daily living, pain, and range of movement at 3, 12 and 52 weeks At one year, patients with normal level of function:
46/53 collar
46/55 mobilisation

Analysis of these dichotomous outcomes is shown in Figure 1 and Table 2. In the two trials where the control was no other treatment, soft collars were no better, with a relative risk for a successful outcome of 1.0 (0.9 to 1.2).

Table 2: Results of trials of soft collar for whiplash, using patients with good outcome at six months or one year

Number of
Percent good outcome
Relative risk
(95% CI)
(95% CI)
No active treatment
1.0 (0.9 to 1.2)
not applicable
0.82 (0.69 to 0.97)
9 (4 to 142)
All trials
0.9 (0.8 to 1.1)
not applicable

Figure 1: Soft collar trials; dark symbols have exercise controls, light symbols no treatment

Exercise and early mobilisation, the control in three trials, were superior to soft collars, producing higher success rates. In these three trials the success rate was only 58% with soft collars, compared with 70% with exercise (Table 2). Using a soft collar instead of exercise would result in a statistically poorer result, with a relative risk of 0.82 (0.69 to 0.97), and a number needed to harm of 9 (4 to 142). For every nine patients with whiplash, one would have a poorer outcome if treated with a soft cervical collar.


Soft collars look like a waste of time. Though Bandolier limited its search to recent studies, older trials and observational studies seem to be in agreement: soft cervical collars probably do less good than giving people exercises.

A shame then that they are still used. A survey of recent Welsh practice [1] showed that half of all grades of medical staff in accident and emergency units would use soft collars for whiplash injuries. Advice on exercise was usually given, always by consultants. Physiotherapy was rarely used, though more often by consultants. Staff with greater experience made more appropriate decisions.

For whiplash injuries, soft collars, for most people, are not worth having, and may do less good than exercise.


  1. AJ Logan, MD Holt. Management of whiplash injuries presenting to accident and emergency departments in Wales. Emergency Medicine Journal 2003 20: 354-355.

previous or next story