Good Outcomes vs Intervention Evidence

At Therapy Partners, Inc we just completed a year of using and implementing FOTO (Focus on Therapeutic Outcomes).  Our sample size is growing steadily - by the end of 2006 we should have about 4000 patient episodes entered into the national database. "Outcomes" are a big  buzzword in PT today, and like any buzzword and trend, I think we need to tread with caution.  Many of the clinicians I work with have made the understandable but incorrect leap that "My FOTO scores are good - therefore the specific treatment interventions I used in this case are now  "evidence based"."  On the other hand they state, "I tried the hip manipulation from the Hoeksma et al study on a patient with hip OA and they had a poorer outcome than the national FOTO database.  So now I am not going to use that technique."  When in reality - a good outcome doesn't imply an EBP intervention and a poor outcome may not be the result of choosing an ineffective intervention.  In the case of the patient with hip OA, they may have done even worse without the manipulation intervention.

With this in mind, I was reading an editorial from the March 2005 Australian Journal of Physiotherapy by Herbert et al.  (AJPT has free on-line access after a year).  It is a thought provoking article that reminds us that outcomes measure outcomes not the effectiveness of specific interventions. 

Herbert's quote "Outcomes of interventions and effects of interventions are very different things" rings true. Hahne et al (2004) and Tuttle (2005) are cited by Herbert as examples of manual therapy situations where a patient's outcome to a treatment can be prognostic.  Improvement within- session can be predictive of between-sessions improvement.

As I promote FOTO among my colleagues and the clinic owners within our organization, I recognize a need to make the distinction between the outcomes of interventions and the effects of interventions very clear.  I think the PT profession needs to do this as well. 

How are we going to help the HMOs and third party payers understand this distinction if we don't?

Comments

[Larry Benz]

Outstanding point. Outcomes (whatever that means), EBP (unfortunately misinterpreted to mean many things that it is not), and Patient satisfaction (eeks, I mean patient loyalty) are all significantly different constructs that may in fact overlap but don't necessarily overlap at all. P4P in theory would take all of them into account by an emphasis on "value" purchasing (whatever that means).

[Selena Horner]

Maybe I look at things in a simple manner, but outcomes indicate whether the intitial presentation of the patient in the clinic didn't change, got worse, or improved. Was there a change and if there was, to what degree? Was the change because of intervention or in spite of intervention... no one knows. In some cases, I would argue, that for the situations that the "wait and see" philosophy led to poorer outcomes than intervention probably did have some effect.

I would think that logically the higher the frequency that the clinician is able to appropriately classify or subgroup patients and provide interventions with known effectiveness that there would be a higher probability of positive change and better outcomes.

Outcomes are a nice snapshot quantifying a clinician's experience. If one had a clinician that seemed to substantially reduce disability and pain in a shorter than average amount of time consistently better than the national average and all your staff, I would think it would be a good idea to do a more indepth analysis of that therapist's performance.

I don't think it is useful to look at outcomes on a patient by patient basis, but instead through groupings of like patients. To toss out proven effective interventions because one patient had a poorer outcome than anticipated is an erroneous judgment, in my opinion.

When one discusses a national database and comparing outcomes, well, the definition of the classification of patients needs to be standardized and interpreted consistently among all the clinicians utilizing the database. I may have a patient that has hip and leg pain which is what was coded with ICD-9 codes, but what if what I am treating is actually a radiculopathy? What if I didn't change the ICD-9 codes to reflect radiculopathy versus joint pain because I'm working in a referral situation and it is a whole lot easier to just treat versus argue about a diagnosis code? Obviously, if there are inconsistencies in coding/classifying then to compare outcomes to a national database may have some flaws.

[Sean]

I spoke to a salesperson for both FOTO and a couple other systems. It is my understanding that in the PTOS version of outcomes software you cannot pick and choose which patient data gets sent. But in FOTO you can; therefore it opens the door for manipulating your outcomes data to look better than it really is. Obviously it was the PTOS salesperson who said this. Any ideas if it is true?

[jwmatheson]

Sean,

I wanted to clarify my post with Judy Holder at FOTO before posting my reply. Please see Judy's answer below (she asked me to post this reply):

"The main difference here, I believe, is that PTOS is used for billing - so of course every patient is included in the system. Unless FOTO is able to interface with billing and / or documentation systems, we just do not have access to information on how many patients a clinic actually treats. We are actively working with RMT (Rehab Management Tools) and will be exhibiting with them at PPS - with that compatibility comes the information on actual patient numbers. We have also had conversations with ChartLinks, ReDoc, and PTOS about developing a compatibility.

In the instances where payers are asking for outcomes, we also benefit again because then the population treated by the clinic is identifiable. With HealthPartners, for instance, outcomes data must be submitted for all back patients. FOTO has created an electronic means of gathering and sending that data, but we have also built in a non-participation audit because a reason must be provided (language barrier, cognitive deficit, etc) if the surveys are not obtained. Again, the missing link for FOTO is when there is not a way to identify exactly how many patients a clinic treated. As we move towards payers requiring outcomes, that issue will be taken care of because the provider will have to include all patients in some fashion or not get paid."

I would love to discuss further - comparing FOTO and PTOS. FOTO mainly measures functional outcomes - effectiveness and efficiency of treatment compared to our risk-adjusted national aggregate. To the best of my knowledge, PTOS is just a billing software - does it have a functional outcomes component?

I would love to discuss this further at your convenience. My contact information is below. Have a great day!

Judy Holder
Manager, Provider Relations
FOTO, Inc.
800-482-3686 ext. 38
Fax: 865-450-9484

My purpose of posting this post was to raise the questions of "Why we are collecting outcomes?" and "What exactly the outcomes are telling us."

As we move to P4P - we need to get a grasp on being able to explain and intepret outcomes, patient loyalty,etc.

Thank you for the great comments.

[Rob Wainner]

Excellent point on outcomes and evidence. Consider the Long et al. Directional Preference study. There were a few folks in the opposite matched group (ie. they received flexion exercise when they exhibited centralization with extension)that actually improved and resolved! Likewise, there was a small group in the matched group that did not improve (http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Retrieve&dopt=AbstractPlus&list_uids=15564907&query_hl=3&itool=pubmed_docsum)
Just another reminder that without a common currency for evidence we can draw some very erroneous conclusion when they are based solely on our clinical experience.

Rob