Improving Spine Care Requires a New Research Paradigm

October 14, 2013

I wrote in my last posting that published data document that the current U.S. system of spine care appears to be actually accelerating the development of chronic pain, work disability, narcotic addiction, use of injections and surgery, and guideline-discordant care in general. Sky-rocketing costs tied to increased utilization of unproven invasive procedures and devices have produced little, if any, evidence of any improvement in patient outcomes. The growing use of MRIs generates misdiagnoses while more opioid prescriptions lead to more addiction.

This decline in quality and rise in cost is directly related to ineffective guidelines that stem from 100’s of RCTs of variable quality that all focus on the study population with non-specific (NS) LBP. The very best these RCTs and guidelines can produce is to determine the best treatment for the average patient with this non-specific symptom. Is it any wonder that guidelines conclude that the best treatment is completely non-specific, one-size-fits-all, specifically encouraging everyone to return to activity ASAP and to reassure that this self-care strategy is likely to result in recovery at some point?

Einstein defined insanity as doing the same thing over and over again and expecting different results. If we change nothing and persist in studying NS-LBP rather than making some diagnostic progress, we will learn nothing new.  We will spend another 25 years learning how to best treat the average patient with this NS symptom.

Two published surveys of international LBP researchers, published in 1998 and 2012, both reported the identification and validation of LBP subgroups as the #1 research priority. There is an urgent need to find ways of identifying or characterizing individuals’ actual underlying disorder and treating that, rather than just a NS symptom. That requires identifying diagnostic subgroups.

Kevin Spratt, Ph.D., an eminent spine researcher at Dartmouth Medical Center in New Hampshire, wrote in 2003 about a research strategy intended to identify and validate subgroups. He referred to his paradigm as the “A-D-T-O” research model. A-D-T-O stands for Assessment – Diagnosis – Treatment – Outcome. Just as A-D-T-O represents the conventional order of clinical care of individual patients, Spratt wrote that it should also be the order for investigating and identifying the best treatments for clinical subgroups.

He points out that these four A-D-T-O pillars are connected by three research links. Establishing the A-D link is where the research effort must begin.  This fundamental link requires conducting inter-examiner reliability studies focused on the assessment process-of-interest in an effort to distinguish between those who are members of a diagnostic subgroup and those that aren’t.  Any further research targeting a subgroup cannot be justified if its members and non-members cannot first be reliably identified.

Only after this A-D link is established can the D-T link utilize observational cohort studies focused on members of this subgroup to determine if improved outcomes can be produced with a single treatment. Can data be generated that indicates that this subgroup has a favorable prognosis if treated in a specific way or ways?

Once the A-D and D-T links have been established, the third and final link seeks to determine the best treatment for this reliably identified subgroup. Only now are randomized clinical trials (RCTs) justified to compare treatments previously identified as promising as a result of the prior D-T studies. Substantially improved outcomes with any treatment are required, i.e. the only way, to validate that subgroup.

Any subgroup validated by outcomes can now be subtracted from that very large NS-LBP subgroup.  In the case of one very large subgroup that has substantial evidence in all three of the A-D-T-O links, this large NS-LBP black box is now significantly diminished.  That subgroup is the one whose members’ pain can be centralized/abolished with a directional preference. This subgroup carries a ‘patho-mechanical’ diagnosis of ‘reducible derangement’ where a patho-anatomic diagnosis is not always possible. I’ve talked about that subgroup a great deal in prior blog posts and will be doing so soon again.  Most spine stakeholders are unaware of this evidence, this subgroup, and this A-D-T-O research paradigm.

This A-D-T-O research model is simple, straight-forward, and focuses on moving spine care toward patient-specific treatments. Because it is either unfamiliar to most LBP clinical researchers, or is viewed as too disruptive from the current research convention of conducting high-quality RCTs that target the best treatment for NS-LBP, it is not in wide use yet. Studies that do comply with this new research paradigm have consequently been ignored by clinical guideline panels that are fixated on the high value of RCTs with little regard for the limitations of continuing to focus on NS-LBP.

The A-D-T-O model should be the basis for future clinical research for any disorder, particularly one characterized by a non-specific symptom. Health care clearly needs to shift its attention toward individualized care and that requires a shift in the overall research paradigm.  We should not waste more time and money conducting RCTs of NS-LBP that, over the past several decades, have produced only one-size-fits-all non-specific care that is not improving either the quality or cost of spine care.

I welcome feedback and questions.  Please be sure to click on “Like” if this was useful and enlightening, or “Leave a Comment” by clicking on “Comment” below.  Have a wonderful week everyone.

One Response to “Improving Spine Care Requires a New Research Paradigm”


  1. […] of having little insight into and no solution for these non-specific symptoms, there is now extensive documentation (2) that validates the importance of DP.  This documentation reports […]


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 )

Google photo

You are commenting using your Google 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 )

Connecting to %s

%d bloggers like this: