A Pain in the Neck: A Critical Review on the Efficacy and Risks of Cervical Spinal Collars in the Prehospital Environment

Sara photo

This essay was written by Sara Marr-Phillips and is the winning nursing entry in the 2019 Critical Writing Prize. Sara is studying at Greenwich University and was nominated by her lecturer Scott Goudie.

This review critically discussed the evidence for and against the use of cervical collars on trauma patients with a suspected spinal injury in a pre-hospital setting. The efficacy of cervical spinal collars to immobilise patients has been widely debated over the last 30 years. A growing body of evidence suggests that cervical collars may be ineffective at immobilising the cervical spine and increase the risk of raised intercranial pressure, inducement of pain, pressure ulcers, dysphagia, restricted ventilation, increased risk of aspiration, a compromised airway and mortality. To this date no randomised control trials for the use of cervical spinal collars has been carried out. Considering concerns have been ongoing over the last 30 years around the use of cervical collars, the growing body of evidence of associated risks, and little or no evidence validating their use, provides a great deal of justification for starting randomised control trials to assess the
efficacy of using cervical spinal collars in the prehospital setting.

Key words: Cervical collar; Prehospital care; Trauma; Spinal cord injuries; Spinal

This review will critically discuss the evidence for and against the use of cervical collars on trauma patients with a suspected spinal injury in a pre-hospital setting. The efficacy of cervical spinal collars to immobilise patients has been widely debated in recent years (Sundstrøm, et al., 2014). In the UK each year, approximately 1000 people sustain a new spinal cord injury (SCI) and, as of 2016, 40,000 people are living with the long-term disabilities as a result (National Clinical Guideline Centre UK, 2016). Approximately 15% of those with a spinal column fracture or dislocation will also have an SCI (National Clinical Guideline Centre UK, 2016). This makes them relatively rare with cervical spinal injuries accounting for approximately 2% of hospital admissions (Davis, et al., 1993).  For the year 2015-2016, ambulance trusts in the UK collectively spent £441,103 on cervical collars with individual trust expenditure ranging from £804 to £71,990 (Veljanoski, et al., 2017). Due to growing pressure on the national health service (NHS) and the need to be more cost efficient than ever before; money spent on a piece of equipment that is currently controversial due to a growing body of research questioning the need to use such immobilisation techniques, whilst potentially causing further harm, could be an inefficient use of funds.

History of the Cervical Collar
The cervical collar was first implemented in 1967; it come in to use during the Vietnam war to evacuate soldiers with spinal injuries (Rogers, 2017; Arishita, et al., 1989). Patents from the time include that by George Cottrell (Cottrell, 1966) and Glenn Hare (Hare, 1974) which are the precursors to cervical collars used today. They are described as providing traction of the cervical spine and a degree of immobilisation (Cottrell, 1966). The assumptions that have led to spinal immobilisations are: that the injured patients could have an unstable cervical spine injury, movement may cause additional damage, application of a semi-rigid cervical collar will help prevent these movements, and immobilisation as a procedure is ‘relatively harmless’ so can be applied ‘as a precaution’ (Benger & Blackham, 2009). This perceived lack of risk and application as a failsafe will be further discussed and how it ties to current recommendations.

Current Recommendations 
Ambulance trusts in the UK follow the Joint Royal Colleges Ambulance Liaison
Committee (JRCALC) UK ambulance services clinical practice guidelines, National Institute for Health and Care Excellence (NICE) guidelines and local trust policies in their decision to immobilise patients (JRCALC, 2016; JRCALC, 2017; National Clinical Guideline Centre UK, 2016). The JRCALC state that patients, with the possibility of a spinal injury indicated using the JRCALC (2017) algorithm, should have the whole spine immobilised either by manual immobilisation or with collar, head blocks and spinal support (JRCALC, 2017). Additionally, NICE guidelines recommend protection of the spinal cord from point of injury to avoid secondary SCI using devices such as cervical collars to immobilise the spine and prevent movement (National Clinical Guideline Centre UK, 2016). It is, however, important to note that in the most recent edition of the JRCALC and NICE guidelines more emphasis has been put on defining cases that would be unsuitable for immobilisation with a collar allowing clinicians to use their discretion (JRCALC, 2017; National Clinical Guideline Centre UK, 2016).  In addition to UK based guidelines, the American College of Surgeons (ACS) and the PreHospital Trauma Life Support (PHTLS) guidelines also advocated the use of semi rigid collars as a prioritised procedure in trauma patients (ACS Committee on Trauma, 2012; PHTLS Committee of the National Association of Emergency Medical Technicians in Cooperation with the Committee on Trauma of the ACS, 2010). These guidelines have been adopted by over 50 countries worldwide (Sundstrøm, et al., 2014). With so many countries invested in the use of
cervical collars it is more important now to consider how effective they are and if there are any risks not previously considered.

Cervical Collars Efficacy
Multiple reviews have looked at the evidence in practice for the effectiveness of cervical collars at reducing movement in the cervical spine since the Cochrane review of evidence in 2001 (Benger & Blackham, 2009; Sundstrøm, et al., 2014; Kwan, et al., 2001). Due to a decrease in incidences and death from spinal injuries around the same time cervical collars and immobilisation were implemented in the 1970’s-80’s, assumptions were made that these changes in mortality were due to their implementation (Kornhall, et al., 2017; ACS Committee on Trauma, 2012). They have been used for more than 30 years in countries all over the world, however, there has been limited evidence to support their use (Sundstrøm, et al., 2014; Benger & Blackham, 2009; Kwan, et al., 2001). This may be due to the quick implementation during the Vietnam war for a way to save lives where time was not available to adequately test new equipment. Moreover, there has been a growing amount of evidence for the adverse effects of cervical collars (Benger & Blackham, 2009).  Due to a growing interest in the mid to late 90’s, a Cochrane report was carried out in 2001 to quantify the effect of different immobilisation methods in trauma patients had on spinal stability, neurological disability, adverse effects, and mortality (Kwan, et al., 2001). They identified 4453 eligible reports, however, not a single trial met the inclusion criteria of being a randomised control trial comparing immobilisation strategies in trauma patients with suspected spinal cord injury (Kwan, et al., 2001). This is a remarkable number of reports to not meet the inclusion criteria and speaks volumes around the style of research that has been carried out so far. This identifies a strong need for development of randomised control trials in this area, however, ethical consideration remains at the fore front when considering these types of trials on patients. To gain ethical approval for studies of this kind, where there is the potential of harm to patients, all other avenues of research showing the lack of efficacy in cervical collars and their risks must be exhausted.

One of the most important studies was carried out by Hauswald et al. (1998). They carried out a five-year retrospective chart review in two hospitals to look at the effect of spinal immobilisation on neurological outcome in blunt trauma patients. One was the University of Malaya in Malaysia and the other at the University of New Mexico, Unites States of America. All the patients taken to the U.S. hospital had their spines immobilised at the site of injury, whereas, none that were taken to the Malaysian hospital were immobilised (Hauswald, et al., 1998). These two hospitals had similar radiological, resuscitative, and surgical abilities. This study found a less than 2% chance that the immobilisation had any beneficial effect on the neurological outcome of these patients (Hauswald, et al., 1998). This would suggest that the perceived risk associated with moving a patient with a potential spinal injury without immobilisation may be unfounded. A number of studies have set out to investigate the range of movement possible whilst wearing a cervical collar in non-clinical participants to assess whether they provide the level of immobilisation that their past reputation has been built on (Sundstrøm, et al., 2014). Ben-Galium et al. (2010) carried out a study using nine fresh human cadavers with simulated unstable cervical injuries. They found that, although the cervical collars did not cause the initial injury, they caused significant separation between the vertebras after application. Cervical collars have been suggested to prevent secondary injury, however, this study provides evidence that the cervical collars themselves may be causing some degree of secondary injuries. Although this is a
small study on cadavers, this model has been seen to be indistinguishable from asymptomatic live participants (Ben-Galium, et al., 2010).

In elderly patients, due to the mechanism of injury and possible pre-existing degenerative spinal diseases, cervical spinal fractures are more prevalent (Peck, et al., 2018). Rao et al. (2016) carried out a study on patients that were over 65 and had suffer major trauma, using CT scans to measure the patients chin-brow horizontal angles whilst lying flat (Rao, et al., 2016). The expectation that if ‘neutral alignment’ was the same for all patients then the chin-brow horizontal angles would be similar, however, this study found that the angles varied widely (Rao, et al., 2016). This suggest that trying to put the elderly patients into the classic ‘neutral alignment’ would in fact be out of alignment for what is normal to them. This may exacerbate possible adverse effects from a cervical collar whilst being immobilised, increase the chances of further fractures in those with osteoporosis and the development of pressure ulcers (Peck, et al., 2018). Although generalisation of these results is problematic due to the small sample size of this study, both Rao, et al. (2016) and Peck, at al. (2018) provide more evidence that collars are not providing the level of protection they have long been praised for providing.  A novel approach measuring the efficacy of methods for immobilising the cervical spine prehospitally was carried out by Rhamatalla et al. (2019). Using a dynamic simulation model, they were able to measure the most effective way of immobilising the cervical spine during transport. They tested 4 measures: a cot alone, cot and cervical collar, long board, collar and head blocks, and vacuum mattress and collar. They found that the long board and vacuum mattress measures were the most effective at immobilising the cervical spine. This study provides a previously unexplored measure of the cervical spine in transport. Future studies should use this style of approach to measure movement across the whole patient journey. It is important to note, however, that this study did not repeat the measures with the collar removed, therefore, it is not possible to say whether the cervical collar was effective or not.

Possible Adverse Effects of Cervical Collars
More recent research has highlighted a number of other risks that have been attributed to rigid collar immobilisation. These include: increased intercranial pressure (ICP), inducement of pain, pressure ulcer (PU), dysphagia, restricted ventilation, increased risk of aspiration, a compromised airway and increased mortality (JRCALC, 2017; Moscote-Salazar, et al., 2018; Tsutsumi, et al., 2018). Although some discomfort is expected with manual handling of patients when being transferred from a prehospital to hospital environment; this growing list of risk highlights some previously unconsidered points to be studied further.

Firstly, it has been suggested that cervical collars raise ICP which can lead to intracranial injury (Mobbs, et al., 2002). Mobbs et al. (2002) carried out a study to investigate changes in ICP after the application of a hard collar in trauma patients with a head injury. ICP was measured without a collar, and at three and five minute intervals after application. Significantly higher ICP was found after the application of the collar. This study provides some insight into the experiences of real trauma patients, however, with such a small sample size, further investigations into the impact of hard collars is required to make meaningful conclusions. Suggestions for this raise in ICP include interference with venous drainage, and pain (Mobbs, et al., 2002; Moscote-Salazar, et al., 2017). Further investigation would be required to elaborate on these results with a larger patient sample size and with a more general trauma rather than specifically studying head injuries. A review of studies researching ICP found that there was not enough data currently to support this hypothesis (Moscote-Salazar, et al., 2017).

Secondly, it has been suggested that cervical collars cause PU’s. Ham et al. (2014) carried out a literature review spanning from 1970 to 2011 searching for studies that have measured PUs in patients having undergone spinal immobilisation. They found that PU’s relating to cervical collars ranged from 6.8% to 38% across different studies (Ham, et al., 2014). PU risk factors include: high pressure and pain from immobilisation, length of time immobilised, high injury severity scores, intensive care unit admission, mechanical ventilation and ICP monitoring (Ham, et al., 2014). Of the 13 studies that met the inclusion requirements; only four were clinical studies. They provided suggestions for prevention or reduction of PU’s which included: early replacement of the collar, collar refit and position change (Ham, et al., 2014). This review supports the need for greater clinical research in understanding the risks of PU’s. 

In regard to a compromised airway, Heath (1994) found that when comparing laryngoscopy of patients wearing a cervical collar and those manually immobilised, 64% of the patients wearing a collar had poor views, and mouth opening was significantly reduced. Although time has passed since this study was carried out, it is still an important study in regard to endotracheal intubation in the trauma patient. This highlights the difficulties in the increased risk of aspiration, compromised airway, and difficulties intubating patients that are wearing a cervical collar. Priority of effective oxygenation of the patient should not be diminished by the need to immobilise the cervical spine. Future studies should also keep this in consideration. 

Moreover, in respect to their effect on mortality, Tsutsumi et al. (2018) carried out a retrospective cohort study from the Japan trauma data bank (JTDB) for patients treated 2004-2015 to look at the effect of spinal immobilisation on traumatic cardiac arrest outcomes. 4313 patients were included in this study with 3307 classed as immobilised (back board and cervical collar) and 1006 as non-immobilised. This study found that the return of spontaneous circulation (ROSC) occurred in 25% of patients that were immobilised compared with 41.9% of non immobilised patients before arriving at the hospital. Furthermore, only 1.8% of the immobilised patients survived to discharge compared with the 3.7% that were not immobilised. This study showed that immobilisation was significantly associated with a lower rate of ROSC by the time of admission and survival to discharge (Tsutsumi, et al., 2018). This could be due to a reduced on scene time as immobilising patients takes a longer period of time to complete, however, there were no significant differences between the groups in the total time taken from arriving on scene to the time to hospital (Tsutsumi, et al., 2018). It can therefore be hypothesised that the collar and board are associated with compromising airway management and raising ICP (Tsutsumi, et al., 2018). 

Although there are some clinical differences between countries to be considered, this is a large study which provides evidence for increased survival rate among the non-immobilised group. More studies could use clinicians’ decisions not to immobilise patients as a way of looking at patient outcomes without actively making the ethical decision to not immobilise by study design. This comes with its own limitations as we would need to account for reasons why clinicians made the discussion to not immobilise in order to make valid conclusions from the results and ascertaining causation. It is important to note that even with this increased survival rate, the number of patients surviving to discharge from a traumatic cardiac arrest is still extremely low.
Manual In-line Stabilisation and Lateral Trauma Position
Prasarn et al. (2012) found significant movement during the application and removal of cervical collars. It is possible that the use of cervical collars to immobilise patients is helpful in reducing the risk of secondary injury, however, the process of applying the collar may result in the significant increase in separation noted by Ben-Galium, et al. (2010).  MILS (Manual in-line stabilisation) could be used to negate this risk (Prasarn, et al., 2012). MILS involve holding the head in-line with the attendants’ hands, providing immobilisation (Sundstrøm, et al., 2014). If collars are to continue being used, they should be applied and removed using MILS only by trained professionals, moreover, there would be a need for practice in this skill in order to maintain the high level of care required during the application and removal process (Prasarn, et al., 2012).

Some researchers have considered the use of MILS and lateral trauma position (LTP) to reduce or remove the need for cervical collars altogether. By holding the patients’ head in-line manually the need for collars and blocks are removed whilst still providing a level of immobilisation. The LTP is a variation on the recovery position whereby the patient is log rolled on to their side with their top leg bent at the knee at a 90-degree angle (Kornhall, et al., 2017). This can be achieved with just 2 people (Kornhall, et al., 2017). As with cervical collars, there is limited evidence for their use but there is a growing body of research and support for using these alternative methods of immobilisation in prehospital settings (Kwan, et al., 2001; Kornhall, et al.,

The use of cervical collars has been based off historical rhetoric rather than scientific evidence (Kornhall, et al., 2017). To date, the existing evidence for their use is weak and, moreover, there is evidence to suggest they do the patient more harm (Kwan, et al., 2001). Current reviews suggest either not using cervical collars in the first place or, if using them, removing them at the earliest possible moment once in hospital (Sundstrøm, et al., 2014), although it is recommended that MILS is used to support the application and removal (Prasarn, et al., 2012). Benger and Blackham (2009) go further to recommend that alert and stable trauma patients do not require immobilisation and the patient should choose a comfortable position to lie in. Their routine use comes at a price that could be reduced if they were used only in the most necessary of cases rather than just as a precaution (Veljanoski, et al., 2017) or the removal of this procedure completely (Hauswald, et al., 1998). There is some evidence for the use of MILS and LTP instead, however, these also need further research as to their efficacy (Kwan, et al., 2001). Guidelines should encourage pre-hospital staff to move away from immobilising a patient ‘as a precaution’ routinely and support techniques that have good quality evidence supporting their efficacy. 

The Cochrane review by Kwan et al. (2001) highlights a lack of randomised control trials in the use of cervical spinal collars. To this date no study has been carried out to provide this insight. Considering concerns have been ongoing over the last 30 years around the use of cervical collars, the growing body of evidence of associated risks and little or no evidence validating their use provides a great deal of justification for starting randomised control trials to assess the efficacy of using cervical spinal collars in the prehospital setting.


American College Surgeons Committee on Trauma, (2012). Advanced Trauma Life Support (ATLS) Student Course Manual. 9th ed. Chicago(IL): Americal College of Surgeons.

Arishita, G. I., Vayer, J. J. & Bellamy , R. F., (1989). Cervical Spine Immobilization of Penetrating Neck Wounds in a Hostile Environment. The Journal of Trauma, 29(3), pp. 332-337.

Ben-Galium, P. et al., (2010). Extrication Collars Can Result in Abnormal Separation Between Vertebrae in the Presence of a Dissociative Injury. Journal of Trauma, 69(2), pp. 447-450.

Benger, J. & Blackham, J., (2009). Why Do We Put Cervical Collars on Conscious Trauma Patients?. Scandinavian Journal of Trauma, Resuscitation and Emergency Medicine, 17(44).

Cottrell, G. W., (1966). Cervical Collar. Portland, Oreg., Patent No. 3285244.

Davis, J. W., Phreaner, D. L., Hoyt, D. B. & Mackersie, R. C., (1993). The Etiology of Missed Cervical Spine Injuries. The Journal of Trauma: Injury, Infection, and Critical Care, 34(3), pp. 342-346.

Ham, W., Schookhoven, L., Schuurmans, M. J. & Leenen, L. P. H., (2014). Pressure Ulcers from Spinal Immobilization in Trauma patients: A Systematic Review. Journal of Trauma Acute Car Surgery, Volume 76, pp. 1131-1141.

Hare, G. F., (1974). Cervical Collar. Encinitas, Calif., Patent No. 3850164. Hauswald, M., Ong, G., Tandberg, D. & Omar, Z., (1998). Out-of-hospital Spinal Immobilization: Its Effect of Neurologic Injury. Academic Emergency Medicine, 5(3), pp. 214-219.

Heath, K. J., (1994). The Effect on Laryngoscopy of Different Cervical Spine Immobilization Techniques. Anaesthesia, 49(10), pp. 843-845.

Joint Royal Colleges Ambulance Liaison Committee, (2016). UK Ambulance Services Clinical Practice Guidelines. Bridgewater: Class Professional.

Joint Royal Colleges Ambulance Liaison Committee, (2017). Clinical Practice Supplementary Guidelines. Bridgewater: Class Professional Publishing.

Kornhall, D. K. et al., (2017). The Norweigian Guidelines for the Pre-hospital Management of Adults with Potential Spinal Injury. Scandinavian Journal of Trauma, Resuscitation and Emergency Medicine, 25(2).

Kwan, I., Bunn, F. & Roberts, I. G., (2001). Spinal Immobilisation for Trauma Patients. The Cochrane Library, Issue 2.

Mobbs, R. J., Stoodley, M. A. & Fuller, J., (2002). Effect of Cervical Hard Collar on Intracranial Pressure After Head Injury. ANZ Journal of Surgery, Volume 72, pp. 389-391.

Moscote-Salazar, L. R., Godoy, D. A., Agrawal, A. & Rubiano, A. M., (2017). Effect of Cervical Collars on Intracranial Pressure in Patients with Head Neurotrauma. Journal of Emergency Practice and Trauma, 4(1), pp. 1-2.

National Clinical Guideline Centre UK, (2016). Spinal Injury: Assessment and Initial Management .

Peck, G. E., Shipway, D. J. H., Tsang, K. & Fertheman, M., (2018). Cervical Spine Immobilisation in the Elderly: A Literature Review. British Journal of Neurosurgery.

Prasarn, M. L. et al., (2012). Motion Generated in the Unstable Cervical Spine During the Application and Removal of Cervical Immobilisation Collars. Journal of Trauma Acute Care Surgery, 72(6), pp. 1609-1603.

Prehospital Trauma Life Support Committee of the National Association of Emergency Medical Technicians in Cooperation with the Committee on Trauma of the American College of Surgeons, (2010).

Prehospital Trauma Life Support (PHTLS). 7th ed. Burlington(MA): Jones & Bartlett Learning. Rao, P. J. et al., (2016). Cervical Spine Immobilisation in the elderly population. Journal of Spine Surgery, 2(1), pp. 41-46.

Rhamatalla, S. et al., (2019). Comparing the Efficacy of Methods for Immobilizing the Cervical Spine. Spine, 44(1), pp. 32-40. Rogers, L., (2017). No Place for the Rigid Cervical Collar in Pre-hospital Care. International Paramedic Practice, 7(1).

Sundstrøm, T. et al., (2014). Prehospital Use of Cervical Collars in Trauma Patients: A Critical Review. Journal of Neurotrauma, 31(6), pp. 531-540.

Tsutsumi, Y. et al., (2018). Association Between Spinal Immobilization and Survival at Discharge for On-scene Blunt Traumatic Cardiac Arrest: A Nationwide Retrospective Cohort Study. Injury, 49(1), pp. 124-129.

Veljanoski, D., Grier, G. & Wilson, M. H., (2017). Counting the Cost of Cervical Collars. Prehospital and Disaster Medicine , 32(6), p. 701.