Analysis in rehabilitation

To analyse something is to “examine (something) methodically and in detail, typically in order to explain and interpret it“. [Oxford English Dictionary]. When I was learning how to make a medical diagnosis of a disease, we were taught a method to collect the information and use it to reach a conclusion on the type of disease and where it was. The process was based on a biomedical framework of human anatomy and physiology coupled with a systematic approach to pathology. It worked well. I was not taught any rehabilitation; it was not on the medical curriculum, nor was there any undertaken in hospitals. Further, I have not knowingly read about a methodological approach to the analysis of rehabilitation problems. This page considers an approach and links to further pages that focus on specific aspects of analysis. Its starting point is that the biopsychosocial model of illness, illustrated in this graphic here, must be the analytic framework.


You will be analysing information from the moment you are first aware of a patient, even down to considering the implications of a telephone call being made before writing a letter (for example). No part of the problem-solving process occurs in isolation. Nonetheless, most people will set aside time to review and analyse available information more formally in more complex cases. This page considers how you can undertake a formal analysis systematically.

Reaching a medical diagnosis of a disease is relatively straightforward because it is a closed problem with a single answer. The analysis has a single goal, and one can define success. Rehabilitation problems are, in contrast, open, and there is no single answer. One is seeking a holistic understanding of the interrelationships between different phenomena. The clinician can break the problem down into a series of closed questions, but the number may grow too large to resolve.

The complexity of the situation poses a second, insurmountable difficulty. Few interrelationships are fixed. For example, the relationship between the strength of the quadriceps muscle and gait speed will differ in different conditions and between people. It may sometimes be the limiting factor but often is not.

A holistic understanding of a complex model with unpredictable and (mathematically) chaotic interrelationships is a challenge that precludes any straightforward analytic method. Neural networks cannot analyse the situation because they learn by being trained through achieving the correct answer, and there is no right answer in an open problem.

However, rehabilitation clinicians successfully analyse rehabilitation problems every day, demonstrating that it is possible. The challenge is to understand how rehabilitation experts who are good at analysis undertake it. It is pretty likely that many do not know because they have not considered the matter.

Previous work

I am not aware of any systematic studies on how rehabilitation experts undertake analysis of clinical problems. However, when researching for this page, I came across an article by Steiner et al. on “Use of the ICF Model as a Clinical Problem-Solving Tool in Physical Therapy and Rehabilitation Medicine“. (here)

Within this article, the authors refer to a “Rehabilitation Problem-Solving Form” (RPS-Form) that is illustrated in figure 6 of a report by Eberhardt and Greiner, “Implementation of the RPS-Form in two centres in Indonesia and Nepal“. (here) The form is simply six boxes covering impairment, activities and social participation as perceived by the patient and family on the one hand and the health professional on the other hand. It also included environmental and personal factors.

A Google search on “How do rehabilitation professionals analyse rehabilitation problems.” reported 167,000 results, but the first page had no other relevant papers.

My conclusions are that:

  • there has been very little research into how rehabilitation experts go about analysing clinical problems, a significant contrast to the research into the medical diagnostic process;
  • at least one other group agrees that the biopsychosocial model of illness is an appropriate framework for the analysis

A systematic approach

My suggested approach is based on the biopsychosocial model of illness. It consists of a series of questions applied to each of the patient-centred domains within the model: pathology, impairment, activities, and social participation. You may approach the fields in any order, but it is easiest and most natural to start with activities, working back through impairment and pathology before ending with social participation. The questions are all concerned with consistency in one way or another.

Are relationships observed consistent, one with another?
The first set of questions concerns consistency. The essential question is, “Are the patient’s preserved and limited or lost phenomena in this domain consistent with the known preserved, limited or lost phenomena in each of the other three domains?” For example, is the inability to walk consistent with a diagnosis of a right middle cerebral artery area stroke two days ago? The answer would usually be yes. But if the diagnosis were of a minor lacunar stroke two years ago, the answer would be no.

As another example, is the observed and reported dependence on others for dressing compatible with a diagnosis of mild cardiac angina, whereby the patient can walk up and down stairs without difficulty or pain? Probably not! However, a further assessment might reveal an episode of cerebral hypoxia causing quite marked apraxia.

If there are areas where the relationship appears different from that expected, then the assessor needs to look further, investigating other factors that might account for the apparent inconsistency. The order does not matter, except it is best to start with the most probable other explanation.

Thus one might check the medical diagnosis. Does the person have an additional disease? Is the primary diagnosis incorrect? Alternatively, are there other impairments present that have not been identified? Anxiety or depression are common additional impairments that cause inconsistency. Alternatively, are there contextual factors that account for the unexpected finding? One common contextual explanation lies within ‘personal context” – expectations, beliefs, and lack of motivation are common reasons for inconsistency. I will emphasise that one should investigate both surprisingly poor performance and surprisingly good performance.

Is the relationship stronger or weaker than expected?
This question differs from the first question mainly in the nature and degree of inconsistency. The first question is categorical. If the patient falls into this category on that measure, they cannot be in the observed category on a second measure. In contrast, this question is, “Given the extent of limitation on this measure, is the performance on a second measure what you would anticipate or is it significantly better or worse?

For example, one might observe that someone has arm strength that is 90% normal yet has quite markedly reduced arm function when feeding himself. As before, one needs to look for an explanation, such as fatigue, apraxia, or fear of choking.

Is the change seen what I would expect?
The second variation on consistency concerns recovery or deterioration in conditions where improvement or deterioration is reasonably likely to occur, such as soon after a stroke or in someone with motor neurone disease (amyotrophic lateral sclerosis). Although there are considerable individual differences in the trajectory of change, one usually expects even if only based on change up to some point.

If the rate of change or extent of change falls outside the expected, the clinician should always consider why. As the underlying pathology primarily determines expected changes, the first step should be to reconsider the medical diagnosis. The other likely explanation is the patient’s expectation and, associated but different, their emotional state.

Curiosity – they key clinical attribute.

This page and the following four pages set out the analysis as a structured set of questions.

However, the best analytic approach is to have an insatiable curiosity about everything. From first contact to final discharge, the clinician should be curious at all times. Not in a nosey way, but in a curious way. The clinician should always think and should usually ask, why is that? For example, why did the father not mention the circumstances leading to his daughter’s cardiac arrest? [he may have made a mistake in wiring a bathroom light] For example, why did the family restrict their mother’s activities after being diagnosed with a transient diagnosis even though she had been well for ten weeks after the event? [The neurologist called it ‘a warning stroke’, and a warning requires someone to do something.]

This curiosity should focus on inconsistencies, things that have not been observed or reported, when they would be expected (The dog that did not bark), unexplained gaps in the information given, occasional unexpected comments etc. When curiosity is piqued, the clinician should ask further questions.

Curiosity often generates hypotheses, and all practical analysis depends upon the clinician developing and testing hypotheses. Therefore, when you are analysing a clinical conundrum, you should always start with some idea about the problems, how they arise, and what can be done. You should use the data available should be used to test hypotheses. The intention is not to prove your hypothesis correct – that is not possible; the intention is to disprove other plausible hypotheses.

Further pages.

Four pages consider the analysis in more detail. Each explores an analysis of the four clinical, patient-centred domains of the biopsychosocial model in more detail:

Pathology (disease)

This comes first for several reasons. It is part of a logical order, working from inside the body out into the whole world. Disease not only can be cured, sometimes, but it also determines (a) what problems are more or less likely to arise and (b) the ‘prognostic field’ for the patient’s problems. It is also placed here to emphasise that reviewing disease is as important as reviewing other domains, no more but also no less. To look, click here.

Impairment (symptoms and signs)

This is, perhaps, the most complex part of the analysis. There are three different concepts included within the word impairment when it is used within rehabilitation:
(a) structural loss,
(b) use of a word to represent a construct, and
(c) patient experience.

Impairment is
(a) an essential bridge between pathology and disability,
(b) a representation of the clinician’s interpretation of observations, and
(c) a representation of the patient’s experience.

To explore further, click here.

Disability (activities)

Disability is the focal point of the rehabilitation process, and analysing disability is a central skill – but not the only one. It is the first step on what may be a short and easy journey or a long and complex one. If the analysis is flawed or incomplete, the journey may be a cul-de-sac and waste the patient’s time, and healthcare resources. Though, at first sight, the consequence of an incorrect ‘diagnosis’ (understanding) of disability may appear less severe than in the case of an incorrect diagnosis of disease, an incorrect disease diagnosis will rapidly become evident but an incorrect rehabilitation diagnosis may only become evident slowly. To read more, click here.

Social functioning (participation)

This area is the most difficult to specify, categorise, and measure but it is also the most important area from a patient’s perspective. Moreover, it is likely that a failure to consider social functioning during rehabilitation will not only lead to a less effective and less efficient process but will also be associated with greater long-term societal costs (health and social services combined). Patients who have an inadequate social function, with too few networks and relationships (in their eyes), are more likely to be less healthy and use more healthcare resources. To read more on the analysis, click here.


On this page, I have discussed how one might analyse the data collected during an assessment to understand the patient’s problems and how one might solve them. It leads to the formulation. I have suggested a structured method, but, at the same time, I have emphasised the importance of curiosity, which should focus on areas of inconsistency and areas where you cannot easily explain what you observe. I have also suggested you need to be alert to what you do not observe and to incidental comments made by families and patients.

Scroll to Top