hrfhome.com

The BPS model is now over 40 years old and multiple concerns and considerations regarding its application have arisen during that time. Bolten & Gillett best summarized some fundamental concerns associated with the BPS model in a recent review.⁹

“Engel’s biopsychosocial model has long been criticized for having various kinds of limitations, along with suggestions for improvements. Increasingly, however, there have been more radical criticisms. Such radical criticisms are of two main types: first, that the model lacks specific content, is too general and vague; and second, that it lacks scientific validity and philosophical coherence. Given the popularity of the biopsychosocial model and its presumed status as an overarching framework for medicine and healthcare, such radical criticisms signal significant underlying theory problems.”

The Human Rehabilitation Framework (HRF) was specifically developed with an acute awareness of the difficulties presented by the BPS model in both its understanding and application. The HRF development process began by first addressing the challenges of philosophical coherence that arises in understanding and applying the BPS model. A lack of coherence arises when there is a lack of consistent adherence to a defined philosophy of science built on a clearly defined worldview and its truth criterion assumptions.

In developing the HRF, a pragmatic scientific philosophy of functional contextualism was chosen and committed to. Functional contextualism defines its truth criterion as “successful working.” This truth criterion was derived from the contextualism worldview defined in Stephen C. Pepper’s book, World Hypotheses: A Study in Evidence.¹⁰ This book presents the argument that humans possess limited ways of “cognitizing” despite the appearance of countless numbers of “schools of thought” traditionally presented in the eclectic tradition of philosophy. In his book, he discusses 7-8 (depending on how you read it) ways of seeing the world, but of these, there are only four relatively adequate philosophical theories or ways of creating world hypotheses that are unlimited in scope. These world hypotheses are derived from “root metaphors,” which are common sense, yet fundamental ways of perceiving the world.

Figure 1: Scheme of adequate world hypotheses built on relative theories as defined by Stephn C. Pepper¹⁰

What is particularly helpful about Pepper’s root metaphor theory is that he removes much of the incoherent eclecticism of the tradition of philosophy arguing over ontological and epistemological debates and provides a practical way of looking at philosophy, which is particularly helpful for developing scientific philosophies. To be concise, we have presented a limited synopsis of the four worldviews suggested by Pepper. The root metaphor of contextualism was selected due to its capacity to work with known corroboration and flexibility with unknowns, to a philosophy that supports an HRF theme of developing “confident ambiguity” in clinical practice. In contrast to the two other most commonly represented world hypotheses of mechanism and organicism in application of the BPS model. Mechanism is dependent on high levels of certainty and precision, with truth specifically associated with “specific parts working together to yield specific outcomes.” Mechanism is highly prone to certainty derived combinatorial explosion. Organicism effectively sees truth as everything having meaning in the whole and everything is a part of the whole. Meaning is implied, therefore there is a risk for teleological error. The fourth relatively adequate world hypothesis, formism, focuses on truth in naming. It is the root scientific philosophical lens that led to the development of ICD and DSM diagnoses. These diagnoses are a particular area of problem in the BPS model.

Human pathologic conditions frequently encapsulate multiple concurrent ICD and DSM diagnoses for a single individual.

“…a hypothetical 79-year-old woman with osteoporosis, osteoarthritis, type 2 diabetes, hypertension, and chronic obstructive pulmonary disease should take as many as 19 doses of medication per day. Adherence to five separate sets of clinical practice guidelines for the woman’s five diseases could result in adverse interactions between her medications, or a medication for one disease could exacerbate the symptoms of another. Such guidelines might also make conflicting recommendations for the woman’s care. If she had peripheral neuropathy, guidelines for osteoporosis would recommend that she perform weight-bearing exercise, while guidelines for diabetes would recommend that she avoid such exercise.¹¹ These situations create uncertainty for clinicians and patients as to the best course of action to pursue as they attempt to manage the treatments for multiple conditions.”¹²

ICD and DSM diagnoses yield interventions based on protocols for syndromes. This is a flawed approach to diagnosis and intervention because no two individuals will respond to a set protocol similarly. Multiple factors that make an individual unique must be excluded in the assumptions of a protocol intervention.

“The complexity of the U.S. healthcare system means that patients and clinicians have more information to consider and more decisions to make than ever before. Often, these decisions are neither easy nor straightforward, and they include varying options, trade-offs, benefits, and risks. Further complicating matters, patients often lack the information they need to make decisions. Fewer than half of patients receive clear information on the benefits and trade-offs of the treatments for their condition”^13-15

“As clinicians’ endeavor to provide the best and most appropriate care for their patients, they also struggle with the cognitive complexities inherent in making care decisions.”¹⁶

To address this as well as many other shortcomings of the protocols for a syndromes approach, a process-based approach to diagnosis and provide interventions will be proposed which serves as the clinical foundation for the HRF.

As we attempt to address the shortcomings of the BPS model, three criteria of key qualities arise: scope, precision, and depth. Beginning with scope, a new BPS framework for rehabilitation must apply to a large range of phenomena, in fact, it would need to be universalistic in nature to cover the scope of domains of orthopedics, neurology, pediatrics, geriatrics, and beyond. A functional contextualism philosophy of science ensures a near infinite scope for the application of the BPS model.

Regarding precision, we are most interested in personalized rehabilitation. If we are to be person-first in a BPS model we cannot be so broad as to fall to general heuristics and loose metaphors, nor can we achieve a reductionistic precision without falling into combinatorial explosion. Furthermore, we run into precision challenges attempting to obtain group to individual generalizability by way of ergodic theorem.¹⁷

Finally, it’s important to address the concerns regarding a lack of specific content associated with the BPS model. We propose that existing content can be adapted through adopting a critical lens of scientific depth. Depth is the consistency and coherence that must be maintained between well-established scientific findings at different levels of analysis across multiple domains (IE: physiology, psychology, sociology, etc.). To maintain scientific depth across content, an evolutionary perspective has been selected. To quote Theodosius Dobzhansky, “nothing in biology makes sense except in the light of evolution.”¹⁸ We further expand on this with Niko Tinbergenny’s proposal that biologic behavior must be able to answer 4 central questions:¹⁹

• What is the history (phylogeny) of the behavior?
• What are the developmental explanations (ontogeny) of the behavior?
• What is the function (adaptive value) of the behavior?
• What is the mechanism(s) (causation) underlying the behavior?

Figure 2: Diagram of Tinbergen’s four questions, divided by object of study and level of question. Source: Bio 342 2012 website

While developing the HRF, several questions arose regarding whether the ICF already accounts for what the HRF looks to achieve. The simple answer to this question is they serve two completely different functions. The HRF is a universal rehabilitation framework for problem framing, clinical decision making, and clinical intervention. The HRF was not developed to be exclusive to pain but is able to expand throughout all rehabilitation as it provides practical strategies for all domains and specialties that reside in healthcare. The ICF is a measurement system used to standardize classification of function and disability with the intention of standardizing language between disciplines and professions and to be complementary to the ICD. The ICF does not provide any resources for problem framing, clinical decision making, or intervention. There are benefits to the ICF in terms of standardizing research terminology and the HRF. Future resources toward this end will be provided on this site in the future. Until then, more information about the ICF can be found through the World Health Organization.

Evidence-based practice (EBP) classically follows Gordon Paul’s famous statement of “What treatment, by whom, is most effective for this individual with that specific problem under which set of circumstances, and how does that come about?”²¹ Even the order to the questioning seems to have led to this explosion and predominance of protocol-based approaches that focus on quickly getting to the “treatment” process without adequately framing the problem in advance. The illusion of good clinical decision making via “differential diagnoses” and misinterpretation of Occam’s Razor led to generalized heuristics and algorithmic thinking generating a large volume of the diagnoses that ultimately became the International Classification of Disease (ICD). Effectively clinicians were taught to race to a diagnosis to initiate a protocol. Described in a different way “protocols-for-syndromes” served as precursor for the formistic diagnosis that must be swiftly found so that the clinician can rush to mechanistically “fix” the problem before the person’s meaning and role in the whole is altered in an organicism way. This approach to problem framing breeds eclecticism, and the Engels introduction of the BPS model has only complimented it.

What Paul’s statement also alludes to are the problems of certainty. It is an attempt to develop an algorithm, which is a mathematical effort to obtain certainty. The problem is, to be certain, how much of the problem do you have to scan? ALL OF IT!

Figure 3: Pie graph representing the knowledge needed required to frame the entirety of a problem.      

This is an impossible task. In the attempt to develop algorithmic approaches to problem solving, we immediately are presented with a problem of combinatorial explosion.²² To quote John Vervaeke, it becomes “cognitive suicide”.²³ In clinical practice, we cannot be a machine looking for certainty. Our deductive logic is built on certainty as it is algorithmic, and this is unachievable as we cannot be comprehensively logical for countless reasons.

This problem is particularly obvious when it comes to the concept of attempting to determine a specific diagnosis related to even a “simple” pain problem. The differential diagnosis process reveals immediate combinatorial explosion even when the problem is examined purely from a biophysiological perspective as shown in the example below of attempting to differentially diagnose between two commonly used shoulder pain diagnoses:

Figure 4: Depiction of simple algorithmic reasoning to determine the mechanistic cause of shoulder pain.

In order to effectively frame a problem while minimizing combinatorial explosion, we must find a way to shrink the problem space that we are searching for important clinical information.

Figure 5: Pie graph representing the knowledge required to frame a smaller portion of a problem.      

Early attempts at utilizing algorithms in clinical problem framing in the form of Clinical Prediction Rules (CPRs) have consistently failed to demonstrate clinical value.^24-26 While many proposals for problem framing are still built on algorithms despite this reality, some theorists in rehabilitation have recognized the limitations of problem framing with algorithms and pivoted to general heuristics. These include Clinical Practice Guidelines (CPGs) which also have failed to make any meaningful impact on clinical decision making or outcomes.²⁷ Beyond these more formal attempts at general heuristics. Other attempts have included Mechanical Diagnosis Therapy (MDT), manual therapy’s “Find it and fix it” approaches, reducing complex problems to “stability” and “instability” causes, and even the best broad efforts such as the axiom “calm things down and build things up” are profoundly limited in how they can apply to human problems.

Fundamentally, the problem with heuristics is you can do all these things and still not achieve your goal. Heuristics help because they limit the problem space you need to examine, but it forces you to prejudge what you are going to select as relevant. It is where we get the term “prejudice” from, or better known as, “bias.” These biases can be helpful and even “adaptive,” but they predispose us to deceiving ourselves and our patients.

This brings us to discussing intended effects, side effects, and adverse events. While these terms are very commonly discussed in the pharmacological domain of medicine, it is nearly ignored in rehabilitation. Side effects are often confused with adverse events, as side effects can be therapeutic, and in a Process Based Approach (PBA), this includes transdiagnostic therapeutic effects that may be beneficial to any given problem. Adverse events are as the name implies, harmful, and occur acutely and chronically when associated with rehabilitation. While acute adverse events such as pneumothorax caused by the incidental event of dry needling the thorax is simple to understand, the iatrogenic effects of rehabilitation education resulting in nocebo is more nuanced. Some adverse events will always be unavoidable, but our intention is to reduce the number and severity of these occurrences.

To act without clarification of the nuances increases the risk of adverse events and the number or severity of side effects. There are always going to be side effects for every action we take, and every intervention utilized. If healthcare is supposed to be bound by the Hippocratic oath of “Do no harm,” then we must start there. By adopting a PBA for the HRF, we can assume some level of precision (personalized) specific effects in addition to a larger scope of transdiagnostic therapeutic side effects. Both this precision and scope are inseparable for depth, which is an attempt to reduce the number and the severity of adverse events. In addition, continuous curiosity that must be maintained in a PBA increases the likelihood of identifying the development of adverse side effects.

Returning to the EBP discussion, PBA asked the core question of EBP in a different way: “What core BPS processes should be targeted with this client given the goal in this situation and how can they most efficiently and effectively be changed?”^28,29

With this in mind, we want to introduce PBA as a relatively new way of framing the problem space. Processes of change allow for effectively, dynamic, flexible, porous containers of multiple domains and levels of mechanisms that account for history and context, and have good scientific depth, as a way to flexibility manage combinatorial explosion in a clearly defined space. They allow you, or rather, force you to slow down and frame the problem in recognizing that some “thing(s)” is sustaining/maintaining the problem. In a PBA, these are assumed to be processes of change that are maintaining the problem in some dynamic contextually relevant manner, and that is how it operates now based on its history.

Once we frame the problem, we must commit to some measure of change occurring that is relevant to the problem. After defining the measure and the problem, then, and only then, can we work with it. This is where an intervention is selected. Selecting an intervention has a high degree of flexibility and openness, as nearly anything can be used to shift a well-defined PBA problem frame. However, it would be wise to select an intervention that most precisely fits to the specific areas of a problem frame network by way of the client and the contextual relevance. Once the intervention is chosen and used, we must measure whether change has occurred. Since the client was involved in the measure, they also must agree that change has occurred. Only then can we confirm whether the most relevant processes were selected, and the most adequate intervention was chosen to yield measurable successful progress.

In addition to core philosophical and content problems that have arisen, there are a multitude of practical scalability considerations that must be addressed. Specifically, many early attempts at stand-alone intensive multidisciplinary programs had difficulty with implementation and efficacy.³⁰ Many of these stand-alone centers and programs have several logistic, financial, and implementation challenges. While there has been some evidence of benefit from multidisciplinary pain programs, the current healthcare climate (organizational factors, reimbursement, etc.) and the likelihood of scaling centers to meet the needs of this pain epidemic are small. Some of these shortcomings can be addressed by targeting alternative, more scalable, options to traditional standalone multidisciplinary approaches. This could include a transdisciplinary training model.³¹ Utilizing a transdisciplinary, and perhaps more broadly, a universalism perspective, allows for individual providers to receive cross training with key aspects of pain treatment provided by other disciplines. This would permit clinics with fewer internal disciplines to work within their network and community to allow for the flexible integration of other disciplines on an as needed basis to meet the practical, clinical, and patient needs. However, this only partially addresses the organization’s cultural and reimbursement problems previously identified. The HRF proposes potential solutions for these and other problems, but some of these will not be available in the initial white papers and publications.

“Unfortunately, these problems are not the fault of an isolated professional domain; the self-perpetuating instrumentalism of the health care system is just one example of the ways in which late modern expert systems target problems that have been defined in their terms, with little time for anything else.³² If, however, efforts to alleviate suffering remain merely instrumental they may cause unnecessary suffering, because human beings do not merely need maintenance or repair: they need to be recognized as human beings.”³³

“We need approaches that make it possible to appreciate technology, professional cultures, and organizational systems as means to ends; they are not meaningful in themselves but only insofar as they serve the needs of finite—vulnerable, creative, and unique—human beings.”³³

The future of research in rehabilitation and other biomedical science is likely to see a significant shift to idionomic methodology. This section will be fully developed in subsequent revisions, until then, we reference the work of Mangalam and Kelty-Stephen:³⁴

“Growing awareness among the scientific community about lapses of reproducibility in biomedical sciences, including psychological sciences and neurosciences have inspired recent developments like the Nature series entitled “Challenges in irreproducible research,” the Reproducibility Initiative, a global project intended to identify and reward reproducible research,³⁵ and increased transparency and data sharing practices. Although we applaud these efforts, the focus has mostly been on the fallacies of P-values, small sample size, inaccurate estimation graphics, and reporting biases. A fundamental problem pervasively linked to the lack of reproducibility in human subjects’ research—inherent in standard analytical techniques—that remains to be considered is nonergodicity, the paucity of group-to-individual generalizability. Crucially, despite the expectation that group treatments will inform individual-level interventions or outcomes, human-subjects research may take ergodicity for granted when it should not.”