Spinal surgery better than nonsurgical treatment over 4 years?

So todayI received the Medscape Week in Review email, which included this gem –

I figured that since it was in Spine, it would favor surgery – sure enough, they found out that spinal surgery is better over the long term, because their original analysis did not account for “nonadherence to treatment assignment” which “caused the intent to treat analysis to underestimate the treatment effects.”  That sounded interesting to say the least.  But I could not find details of the non-op care other than it included at least “active [PT], education/counseling with home exercise instruction, and [NSAID’s] if tolerated…. individualized for each patient and tracked prospectively.”  The criteria to be in the study was signs and symtoms of lumbar radiculopathy that persisted for at least 6 weeks and a herniation at the corresponding level and who were surgical candidates. 

My first thought – isn’t it a good research approach to analyze with intent to treat?  Just because a certain number of your subjects jumped ship and went with surgery doesn’t mean you get to change the rules to get better results – it could skew it in your favor.  They did use a Global Hypothesis Test to analyze for this, and a limitation they admitted was that “as-treated analyses presented do not share the strong protection fromconfounding that exists for an intent-to-treat analysis.  However, the as-treated analyses yielded results similar to prior studies and to a well designed, randomized trial by Peul et al.  Another limitation is the heterogeneity of the nonoperative treatment interventions, discussed in our prior papers.”

And what about the non-op treatment?  I went back to the 2006 edition of JAMA (Surgical vs nonoperative treatment for lumbar disk herniation: the spine patient outcomes research trial (SPORT): a randomized trial; Weinstein JN, Tosteson TD, Lurie JD et al.  JAMA 2006; 296(20):2441-2450) where they talked about initial results (no big surprise, they found no difference between the groups).  “A variety of nonoperative treatments were used during the study.  Most patients received education/ counseling (93%) and anti-inflammatory medications (61%)(NSAIDS, COX-2 inhibitors, or oral steroids); 46% received opiates; more than 50% received injections, and 29% were prescribed activity restriction.  Forty-four percent received active PT during the trial; however 67% had received it prior to enrollment.”  So you have surgery, which is very defined both in the study and in the clinic, and you compare it to no less than six different options, three of which are drug options.  No standardization given for the education/counseling component (did one doc recommend a lumbar roll and to lift with the knees, while another recommended stretching regularly and keep a neutral spine?  What caused them to make the recommendations, and is one better than another, even if the patients had the same presentation?), and especially not for the PT.  The original study was flawed in the first place – use the best evidence out there, and standardize your protocol or have an algorithm/ decision tree!  They acknowledged in that paper that there was heterogeneity in their approach, but “given the limited evidence regarding efficacy for most nonoperative treatments for lumbar disk herniation and individual variability in response, creating a limited, fixed protocol for nonoperative treatment was neither clinically feasible nor generalizable.  The nonoperative treatments used were consistent with published guidelines” (which were Acute Low Back Problems in Adults by the Agency for Health Care Policy and Research, 1994, and a chapter on herniated discs in the North American Spine Society Phase III Clinical Guidelines for Multidisciplinary Spine Care Specialists, by the North American Spine Society in 2000).  Not every patient is the same, but you can’t compare apples to oranges, and you need a clinical decision tree or algorithm, not just for studies but for the clinic (somehow it works for their surgery, but it can’t work outside the OR).  I can maybe understand if they started the study in early 2002 why they used the 2000 edition, but to redo your stats while more research has come out favoring PT and using an algorithmic approach seems to be a bit… ignorant?  Wrong?

So why blog about this?  Late last week I worked with two patients who had spine surgery (one a discectomy and one a fusion), and reading a bit about the background (and given my last clinical was with a great outpatient manual therapist), I wondered if it was necessary or worthwhile.  I wish them the best, but if I or my relatives ever have back problems where surgery is suggested, I will ask for or recommend PT, and find the best damn therapist I can that has experience with LBP and has an evidence-based approach.  This evidence doesn’t hold water for me as the original study was poorly designed (IMO) and the secondary analysis of the data is even worse (again, MO).

(thanks to those PT researchers out there and those promoting EBP)


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s