GP vs. PT Treatment for Whiplash

Hello all.  If you haven't seen the attached article from Spine (Education by General Practitioners...),  I am betting you will soon be hearing from someone who has.  It was brought to my attention by a colleague at UF - Thanks, Sandy. I had not seen it mentioned in this blog yet, so figured I would post. 

I don't mean to ruin the ending for you, but the basic message of the study is "Treatment by GP's and PT's were of similar effectiveness".  Long term endpoints seemed to favor GP's, but there were some variation in endpoint comparisons, with most equivalent, and some even favoring PT's.   

In short, I am not too impressed with these findings.  While the authors used a high quality study design (randomized trial), their 2 intervention groups did not differ much at all.  For a true description of the groups I had to go back to the original article (same first author, 2003). 

Here is how the authors distinguish the groups from page 416 of the 2003 article (also attached):

"Contrast between 2 intervention groups (GP versus PT) is created by the profession of the provider (GP versus PT), duration of treatment session (10 minutes versus 30 minutes), and intervention possibilities (education, advice, and advice on graded activity by GP versus educated, advice, graded activity, and exercise therapy by PT)".

In my opinion, there is too much potential for overlap for these intervention groups to differ drastically.  Essentially, they randomized patients to the same treatments, which may go along way in explaining the variation in observed outcome "differences". 

Furthermore, without including a natural history group, we have no idea if either of these treatment groups were better than doing absolutely nothing, which seems to be an important issue in this particular group of patients.

I hope I am wrong, but this study appears to be another example of a poorly designed RCT that adds little/nothing to the knowledge base for how to (or how not to) manage patients with WAD.

Download 00007632-200604010-00003.pdf

Download DescriptionofWhiplashTreatments_2003.pdf

Comments

[meder]

Medical News
http://www.drplace.com

[John Ware]

I would refer all to Twomey and Talyor's article from the JMMT: The whiplash syndrome: pathology and physical treatment, [vol 1. no. 1 (1993), 26-29] for a sensible approach to conservative treatment of the whiplash patient. RCT's like the one Steve sites just lack common sense with respect to the intervention. According to Twomey and Taylor, many, if not all whiplash pt's should be immobilized at least for a few days following injury, then start on ROM exercises progressing to isometrics. They should remain in the collar for several weeks depending on their signs and symptoms. They also warn against manipulation until healing occurs. Their study is an anatomical descriptive one of post-mortem cervical spines-not an RCT, though it makes a heck of a lot of sense to me. Many RCT's seem to lose common sense when they attempt to standardize interventions. There's no such thing as a "standard" whiplash injury, how could there be a standard treatment?

[Jeff Hebert, DC]

Hi John,
I was surprised to read your response regarding the treatment of whiplash. I attempted to find the article you cited, however it appears that the journal is not listed on pubmed. My recollection was such that the overwhelming majority of the literature pointed to early mobilization and the avoidance of collars. I performed a very brief search on pubmed with the words collar and whiplash, restricted to reviews. The first 3 citations are listed below and seem to refute the use of immobilization.

While Twomey's treatment recommendations may be intuitive and sensible, I'm not sure that there is much out there in terms of support in the literature.

Jeff

1: Revel M. Related Articles, Links
[Whiplash injury of the neck from concepts to facts]
Ann Readapt Med Phys. 2003 Apr;46(3):158-70. Review. French.
PMID: 12763647 [PubMed - indexed for MEDLINE]
2: Weinhardt C, Heller KD. Related Articles, Links
[A systemic review of the value of physical therapy in whiplash neck injury]
Z Orthop Ihre Grenzgeb. 2002 Sep-Oct;140(5):499-502. Review. German.
PMID: 12226772 [PubMed - indexed for MEDLINE]
3: Eck JC, Hodges SD, Humphreys SC. Related Articles, Links
Whiplash: a review of a commonly misunderstood injury.
Am J Med. 2001 Jun 1;110(8):651-6. Review.
PMID: 11382374 [PubMed - indexed for MEDLINE]

[Mark Armstrong, PT, DPT, OCS]

There seems to be a need for better RCTs with this population. I know Dr Sterling has done good research on widespread sensory hypersensitivity as a measured construct. Maybe she can comment or describe better studies out there. Also JOSPT has a good neck classification proposal (JOSPT, Nov 2004, Childs, Fritz, Piva, Whitman). Whiplash disorders fall under the classification of pain control--they recommend 'gentle AROM within pain tolerance'.

[Louie Puentedura]

This is a very interesting topic and the discussion so far has been great. I wanted to add information on a PhD study that was conducted by Dhakshinamoorthy Gurumoorthy, at Curtin University, Perth, Western Australia under supervision from Lance Twomey. I'm not sure why Dr. Gurumoorthy's study results have never been published, but his PhD Thesis can be read from the School of Physiotherapy at Curtin University in Perth.

Here is the title and abstract:

A study of neck injury arising from motor vehicle accidents and its clinical management. PhD submitted 1996.

Abtract:
The syndrome commonly referred to as “whiplash injury" resulting from motor vehicle accidents is complex and remains a challenge to clinicians, as is evidenced by the recent report of the Quebec task force on the "whiplash syndrome". The main objective of this prospective randomised study was to evaluate two conservative treatment regimens (early immobilisation-experimental group-1, early active mobilisation experimental group-2) which are based on accepted physiological rationale and then to compare their effectiveness with existing treatment regimens that are commonly practiced (control group) in the management of "whiplash" type of injuries. To this stage, the current study is the only prospective randomised clinical trial of its type conducted with a sufficiently large sample size and over a long study period. The results of the current study clearly demonstrated that the subjects in the immobilised group recovered from their pain-related symptoms and returned to their normal duties sooner than those in the other two treatment groups. In addition to this, those subjects who received the immobilisation regimen did not show adverse effects on either the range of motion or the strength of the neck muscles. Thus, the immobilisation regimen was clearly shown to be the preferred option when compared to the other two treatment methodologies investigated in the current study.
Although the primary interest of the current study was to compare the efficacy of three different treatment regimens, a series of statistical analyses were performed to establish the prognostic significance of several factors associated with "whiplash" injury. This showed that factors such as gender, age, speed of the vehicles involved, paraesthesia, intensity of pain at the time of the initial examination, interscapular pain, blurred vision and difficulty in focusing, all had prognostic value. Similarly, the type of collision, seating position, presence of headache within 24 hours post injury, pre-existing degenerative changes in the cervical spine, loss of lordosis and litigation factors had no prognostic significance. Another major emphasis of the current study has been to concentrate on the pain related symptoms of the neck which are of major concern to "whiplash" subjects and to those clinicians treating them. A paucity of such information is considered to be one of the most notable causes of difficulties encountered in the management of "whiplash" injuries.
As an adjunct to the main study, the morphology of the deep pre- and post vertebral muscles of the neck region using embalmed cadavers and fresh post-mortem specimens was investigated, as the literature is deficient in--this regard. Similarly, a longitudinal study of 45 subjects was also performed using Magnetic Resonance Imaging (MRI) technology. The longitudinal nature of the M.R.I. study provided for the first time an account of the details associated with the progressive pathological changes that occurred in some disc lesions, at defined points in time following a MVA. The observations made from the adjunct studies help develop a better understanding of the pathoanatomy associated with the deep muscles of the neck region and the pathological changes that occur following a traumatic disc lesion as evidenced within 2 weeks, after 3 months and 12 months post- injury. On the basis of the observations made in the current study, a classification of the "whiplash" injury has been proposed for the consideration of clinicians. Similarly, the questionnaire used for data collection in the current study, can be readily modified and utilised in a clinical situation for establishing documentation, planning treatment strategies and for evaluation of the treatment outcomes of "whiplash" type of injuries.

[John Ware]

Jeff,
Ah, the perfect opening for a plug! The Journal of Manual and Manipulative Therapy (JMMT) is the official journal of the American Academy of Orthopedic Manual Physical Therapists (AAOMPT). If your were to join, you, too, would have this excellent tome sent directly to your address of choice (under $100/year-a real bargain!)
In fact, Twomey and Talyor's article does support early "mobilization" in the form of pain free, usually active, ROM exercises. The use of the collar is when the pt is not exercising(not unlike the way PT's treat acute extremity injuries every day). They are not proponents of passive, manual mobilization in the acute and subacute phases of healing. Their exquisite anatomical studies of the extensive damage to the the complex intervertebral articulations and soft tissues following whiplash in my view are prima facie evidence in favor of the use of early immobilization. I had the privilege of having Dr. Twomey speak to our residency class in Gulfport in 1997. He referred to the study sited above by Louie, but I couldn't find my own notes and therefore didn't want to refer to some unkown study by an unnamed author (thanks, Louie).
I looked at abstracts of a couple of the articles you sited. The interventions are attempts at standardizations with an "all or nothing" approach. That is, one group is immobilized for some predetermined, arbitrary period of time, while the other one starts moving, or is moved, early. I agree with Mark that the quality of the RCTs needs improvment in this area.
My question is this: What do you do with a fresh grade 2 ankle sprain? Why would you treat a swollen, bloody, painful neck any less cautiously?
John

[James Elliott]

I have read with great interest the posts within this thread. I might also add but a few thoughts in relation to whiplash associated disorders and current research endeavors.

Michele's most recent manuscript:
Sterling M, Jull G, Kenardy J. Physical and Psychological Factors Maintain long-term Predictive Capacity Post-Whiplash Injury. 2006 Pain-in press.
To summarize the results of the study: WAD subjects with ongoing moderate-severe symptoms (based on NDI score of > 30/100) continue to manifest Motor Impairments (reduced C/sp ROM, increased EMG activation during cranio-cervical flexion), sensory hypersensitivity and elevated levels of psychological distress when compared to those recovered participants (NDI < 8) and those with milder symptoms (NDI 9-29). The latter two groups continued to only demonstrate motor deficits but no signs/symptoms of abnormal pain processing.

Thus, higher initial NDI scores, older age, cold hyperalgesia, and post-traumatic stress remained significant predictors of poor outcome at long-term follow-up (2-3 years).

It is these features that suggest complex mechanisms that contribute to the development of persistent mod/sev symptoms following whiplash injury; these include reduced temperature thresholds, post-traumatic stress and motor deficits and it is occurring in about 20-25% of the population.

As stated in the Scholten-Peeters article, most WAD subjects fall within the category of WAD II (musculoskeletal signs) and most subjects within WAD II will spontaneously resolve. That said, a significant proportion (~ 19-60%-Barnsley et al, 1994) of those subjects will however continue to suffer from persistent pain and disability in the long-term. The question is why?

It should also be mentioned that there is a tremendous amount of variability within the category of WAD II and as such, treating subjects with the category of WAD II as a homogenous group will not likely result in lessening the transition from acute to chronic pain for some patients. I will let Michele divulge further, but there is work currently underway looking into treatment of chronic and acute WAD subjects based on the presence of altered pain processing and psychological distress. More to follow.....

In short, I am not surprised to learn of the results from this most recent RCT. The mean NDI score for the GP group was 18.5 ± 9.1 and 20.3 ± 7.0 for the group receiving physio. At 52 weeks the NDI scores fell by 8.4 ± 7.4 in the GP group and 6.8 ± 6.2 in the physio group. Thus, if we just look at the NDI scores alone, these are subjects who would fall into either the recovered or mild categories. I might argue that these subjects were likely to improve no matter what treatment was provided?? It beg's the question as to how would those subjects with moderate to severe symptoms respond to GP and/or Physio care? Again, more to follow with some results related to that question in the near future.

Lastly (and to add to Louie's post) we have recently completed our MRI research investing muscular changes within the cervical extensors and relating these findings to the abnormal pain processing mechanisms listed by Michele. More to follow there as well.

I am very interested to read more posts pertaining to this topic.

Kind regards
Jim

[Steven George]

Thanks for that information on the study.

Would it be possible for you to post the article once it is available to the general public? Specifically, I am interested in how sensory hypersensitivity was determined.

I work with a group here at UF that uses a fair amount of sensory testing and am always interested in groups that use the testing procedures in clinical populations.

Thanks.

[James Elliott]

Steve,

you bet, I will post the article once available.

Regarding the sensory meaures: we use a Quantitative Sensory Test (QST) that records thermal (cold/hot) temperature thresholds. The equipment is described in detail on Medoc's website. http://www.medoc-web.com/tsa.html

We use a similar machine at UQ made by Somedic

http://www.somedic.com/article.php?15

They both operate on the same method of limits algorithm.

Jim

[Jeff Hebert, DC]

Hi John,

I apologize for not being able to give this topic the time it deserves- I have really enjoyed the discussion.

You make a good point regarding the ankle, however.....While I have seen many whiplash injuries over the years, I have seen very few which were as you describe: swollen and bloody, akin to a grade 2 ankle sprain. These are injuries which would show up on MRI, and frankly do not bode well for the patient's prognosis. For example, with a front end collision, if the pt had torn their supra and interspinous ligaments, I would think that they may be a good candidate for immobilization, I just don't see it very often. I assume you are relating this to the JMMT study? From what I gather (my online JMMT access does not go back that far), a fair number of the subjects had mechanisms of injury which were so severe, that it was fatal. I would not be surprised to see pt's of this populations have such severe c/s injuries as you describe.

Regarding the mob/manip issue in the acute/subacute stage, I would have to disagree as well. The vast majority of research in this area has pointed to these approaches being most effective in the acute stage. IMHO, one may be limiting the effectiveness of manual tx if it were delayed until the chronic stage.

I appreciate the dialogue.

Jeff

[Jeff Hebert, DC]

Jim,

Thank you for the abstract, this is very interesting work. I have some familiarity with QST as we have a unit in the spine clinic where I work. I was under the impression that its use was only valid when evaluating sensory deficits associated with polyneuropathy. Is there reliability/validity data on evaluating sensory changes from a centrally mediated source? Please understand, that by no means do I claim any expertise in this area, just curiosity.

Thanks,

Jeff

[Steve Young]

John and all,

Thank you all for your excellent discussion on this topic. Not to take things off on a tangent, I just wanted to comment on the immobilization of an ankle sprain. Perhaps someone can correct me, but I believe the evidence suggests that mobilization as tolerated is the treatment of choice and not immobilization. The Chocorane Library states that there was no significant benefit from immobilization of the ankle and reccomends functional treatment. Early mobilization of the ankle resulted in earlier return to work and decreased pain when compared to immobilization for 10 days (American Journal of Sports Medicine 94).

Even Grade 3 sprains seem to do better with functional treatment:

Treatment of complete rupture of the lateral ligaments of the ankle: a randomized clinical trial comparing cast immobilization with functional treatment.

Ardevol J, Bolibar I, Belda V, Argilaga S.

Orthopedic Surgery Department, Hospital de Mataro, C. de Cirera s.n. Mataro, 08304 Barcelona, Spain. jardevol@csm.scs.es

This study compared the therapeutic efficacy between cast immobilization and functional treatment of grade III ruptures of the lateral ankle ligaments. Subjects ( n=121) had closed physeal cartilage, age under 35 years, grade III rupture without previous or associated injuries, and practiced regular sports. Patients were randomized into an immobilization group (21 days plaster cast) or a functional one (15 days strapping plus early controlled mobilization). Symptoms (pain, swelling, stiffness, subjective instability), joint laxity, return to preinjury activity (time and level) and rate of reinjury were assessed 3, 6, and 12 months after sprain. Objective joint laxity was related to constitutional laxity, creating a new variable [talar tilt at injury - talar tilt at control]/contralateral talar tilt. The functional group showed significantly earlier and better return to physical activity, fewer symptoms at 3 and 6 months but no intergroup difference at 12 months. Functional treatment also showed better decrease in joint laxity. No intergroup differences were found in the reinjury rate. We conclude that functional treatment is safe, associated with a more rapid recovery, and particularly suitable in athletic populations.

Sorry to go off topic but I thought it was worth commenting on.

Steve

[John Ware]

Steve,
I think your ankle RCT analogy supports my argument and that of Twomey and Taylor. The goal is to provide an optimal environment for healing damaged soft tissues. That should include a period of relative rest commensurate with the level of injury. I am not a proponent of complete immobilization of the ankle (certainly not in a plaster cast for 21 days!) or the neck following significant non-fx trauma. Relative immobilization to approximate the ends of the injured soft tissues to promote healing along with AROM exercise is a reasonable approach. In the ankle, we can control WB with crutches, in the neck I think a collar may be the analogous assistive device. The study by Gurumoorthy (cited above by Louie) supports the use of rigid collars for whiplash injuries for better long term outcomes. By the way, in an, as far as I know, unpublised case series by Barney Poole, PT, MEd, soft cast material was use for up to a week in severe grade II and III ankle sprain with good results. These pt's also performed supervised exercise and graded return to WB.
Jeff,
I think that it has been well established that of the types of injuries that result from an MVA, hyperextension injuries are the most damaging (MacNab 1964, Forsyth 1964, Talyor and Blackwood 1948, Barnes 1948, Toglia 1976, MacNab 1973, Scher 1983, Davis et al 1991). Some of the injuries that have been described in the above are tears to the longus colli/capitis, retropharyngeal hematomas, esophageal hemorrhaging, anterior longitudinal ligament tearing and separation of the disc from the vertebral body. All of these injuries would likely result in a "swollen and bloody" neck, although that would not be apparent on the surface. Furthermore, in my experience, it is not unusual for pt's to have severe symptoms and high NDI scores when they arrive for treatment, but have negative x-rays and no MRI has been ordered. Many of these, I'll grant you, are several weeks post-injury and have what I would consider secondary joint restriction problems in other areas (CT, t-spine, even maybe suboccipital. I would submit that these folks are presenting with secondary joint restriction because they weren't effectively immobilized immediately after the accident, and have found various mechanisms to compensate so they can function ("global" muscle dominance over injured/inhibited "local" muscles).
John

[Steve Jorgensen]

Great discussion. I think that Jeff's point that a collar for the c-spine is analogous to crutches (in that they may be viewed as a assistive devices) is a good one. In as much as this is the case, the collar may provide a degree of control in the acute stage. There is nothing stopping the caregiver removing the collar to begin some gentle active exercise with the patient - and of course plenty of good evidence, as cited above, to support that this be done. Questions: by what stage do we fully remove the use of the collar? How do we determine this?

Steve.

[John Ware]

Steve,
Perhaps Louie can chime in on the guidelines from the Gurumoothy study. I believe the immobilization group was in a collar for up to 6 weeks. I do not recall the criteria for removal of the collar, but I seem to remember there was a range from 2-6 weeks. The original study is hard to come by since it has not been published. The point is that a reasonable period of relative immobilization makes sense along with supervised ROM exercise. Pt's signs and symptoms have always been a good guide to go by where there is a lack of evidence to support a certain protocol. I personally would not begin any end ROM mobs (grade IV-V) from C2-C3 to C6-C7 for at least 6 weeks. I may perform tx mobs at C0-C1,C1-C2 and C-T region prior to this based on signs and symptoms. I also may do some PROM in cardinal planes for the particularly anxious individual who needs a little encouragement to move. I would always err on the side of caution.
The analogy with the ankle is easy to make. However, while you may render them lame, you're not going to cause someone to stroke out by manipulating their ankle after a bad ankle sprain.
John

[Louie Puentedura]

Steve and John,

I have the entire Gurumoothy study in 17 pdf files so if you'd like to read it, I'd be happy to pass it on.
I checked the methodology section and the experimental group (immobilization) was fitted with a hard Philadelphia collar for 4 weeks. The subjects were allowed to remove the collar for showering, general cleansing, dressing, while applying local heat and in the case of men shaving. They were allowed to apply local heat to their neck and shoulder girdle (hot pack, hot water bottle, hot towel) for a duration of 15 minutes 3 times a day.
At the completion of the 4 weeks the immmobilization group were then instructed in active exercises.
Hope this helps.