Clinical Prediction Rules: Where Do They Fit in Our Clinical Practice?
Written collaboratively by Joey Masri PT, DPT, Philip Van Dyke PT, DPT, CSCS, & Vinh Vuong PT, DPT
Every student or novice clinician wants to become a skilled and proficient clinician but sometimes finds it difficult to figure out which methods work best for a specific diagnosis. With a variety of different techniques taught in school and diversity of topics offered in CEU courses, it can be challenging to determine the right approach for each patient… Are clinical prediction rules (CPRs) the solution? In general, clinical prediction rules are used to help clinicians develop an appropriate diagnosis, prognosis and intervention/treatment plan for patients with given conditions. They were developed to assess a risk-to-benefit ratio associated with a condition and a specific treatment method for that condition. In theory, a CPR may help to prevent possible ramifications related to a physical therapy treatment by quantifying the risk level and efficacy of a specific treatment for a specific diagnosis. Instead of looking for quantitative data related to CPRs, however, our decision making is still often based more on qualitative factors.
For physical therapists, CPRs help guide our decision making towards a more definitive route. Rather than treating every patient with generalized low back pain or shoulder pain with the same treatment approach, the CPRs may point us towards specific protocols to perform depending on whether certain qualifications are met by each patient. This may allow patients to react better to treatment and have longer-lasting results after rehabilitation.
It’s important to note that CPRs are intended to be used for patients with a diagnosis for which there are multiple possible treatment approaches. Whereas the treatment of an ACL reconstruction will not vary greatly between patients, diagnoses such as low back pain (LBP) or patellofemoral pain occur for a variety of reasons and do not all respond to the same type of treatment. Further, every patient is unique in their own way and may present differently symptomatically regardless of having the same diagnosis. For example, you may have two patients come in with generalized LBP, but one presents with additional depression and anxiety while the other presents with eagerness to relieve the pain. Due to the different psychosocial factors, the patient with depression and anxiety may not be an appropriate candidate for utilizing CPRs whereas the other patient is. That is up to your clinical decision making.
While the idea of having a guideline for treatment sounds like an attractive concept, clinical prediction rules are far from perfect. Studies validating the use of CPRs are often not performed and are therefore difficult to confidently merge into clinical practice. Beneciuk et al. conducted a systematic review looking at the quality of a number of studies whose primary purpose was to develop CPRs for specific patient conditions. Using three reviewers, they utilized an 18-item criteria for assessing the methodological quality of these studies, and assigned quality scores to each. According to their scale, studies scoring greater than 60% were considered high-quality. It was stated however that these quality scores should not at all be used in lieu of a validation study for these CPRs, which can provide more definitive information on clinical use. All in all, more rigorously designed studies should provide appropriate insight into their usefulness.
CPRs that scored above 60%:
CPRs that scored below 60%:
So, where do CPRs fit in our clinical practice, especially as students or novice clinicians? Before considering the use of a CPR for a specific diagnosis, we need to critically appraise the quality of research to establish the CPR and any validation studies that are available. As we have shown, not all CPRs are created equal. If you determine that the CPR you are considering has sufficient evidence to support its use, go ahead and use it, but don’t let it be a replacement for using your brain and critical reasoning.
We went through (or are currently going through) three years of school to develop our clinical decision making skills (not to say that those skills do not need further development after graduation), so the CPR should be a tool in our toolbox to help with those decisions. If you are treating a patient with LBP that meets four out of the five criteria for manipulation, that CPR comes in handy. What if the patient doesn’t fit cleanly into that sub-classification or into any of the others for LBP? Well, as would seem self-evident, the research shows that the use of the low back pain CPR is not quite as effective in those cases. Whether or not you choose to factor the CPR subclassifications into your decision making, always remember to assess and reassess your patient in order to actually determine whether your treatment choice is effective. If it is effective, that’s great! If it’s not, it’s time to use those decision making skills again.
Although this is a different topic entirely, and we will not be covering this in depth, it is worth mentioning the difference between CPRs and Clinical Practice Guidelines (CPGs). Whereas CPRs attempt to help prescribe a specific type of treatment for patients that meet certain criteria, CPGs summarize and grade the quality of evidence available about a wide variety of diagnoses. If you haven’t checked out the database of CPGs on PTNow, it is worth the time to do so.
Cook C, Brismée JM, Pietrobon R, Sizer P, Hegedus E, Riddle DL. Development of a Quality Checklist Using Delphi Methods for Prescriptive Clinical Prediction Rules: the QUADCPR. J Manipulative Physiol Ther. 2010;33(1):29-41.
- Stanton TR, Hancock MJ, Apeldoorn AT, Wand BM, Fritz JM. What characterizes people who have an unclear classification using a treatment-based classification algorithm for low back pain? A cross-sectional study. Phys Ther. 2013;93(3):345-55.