How Consistent Support Data Strengthens Functional Assessments

The structural weakness that every honest OT will admit to is that functional assessment samples. An assessment visit is a one-hour or two-hour visit with the person on a specific day, in a specific state, and it skews in both directions, the rally for the professional, or the assessment-day dip on a day no different than Tuesday.

The OT triangulates with interview and history, but the interviews sample memory, and memory samples badly. Outside the assessment process, and far richer in most supported participants, is the functional data of the people who watch them live, hours a day, for good and bad weeks. Most assessments never get to longitudinal functional observation, a density of observation that would elude the capabilities of any clinician.

Transforming that observation into assessment-grade information is a design job, and it's one worth doing, because the stakes of functional assessment, plan funding, equipment justification, capacity trajectories, are exactly what a thin evidence base costs the people involved.

Why raw worker notes are not enough

The problem isn't the desire to get it done, it's the format. The standard shift note is a continuity and compliance note, and is more or less a blank form: it notes that something was done, not how it was done, or how much, or under what circumstances. If you tell workers to write more detail, you will get more detail, but it will be the same level of empty detail. The answer is the same as elsewhere in the series: OT-supplied observational language that specifies what to watch, so the workers supply what they see.

The structured observation request

The working tool is a short observation frame that the OT prescribes for the time leading up to an assessment or review, specifically related to the tasks at hand. For every task that is significant for the assessment: what the person did, what help was provided (and what type), prompting, setup, physical support, how this changed over the course of the task, and what the person did without being prompted. Three or four tasks, a simple format that workers can complete in minutes, a defined window, the fortnight before review, not forever, so the data is fresh, not stale. When stated like this, workers produce exactly what assessment needs: "made a sandwich with setup only on 8 out of 10 days, needed physical help with jar lids every time, initiated making tea unprompted twice," functional evidence in a sentence.

What the data does for the assessment

Structured support observations do three things a single formal assessment can't. They correct the sampling error and determine whether the assessment day was representative, inflated or suppressed, which affects the conclusions on capacity. They show variation, which is required by episodic and fluctuating conditions and is not necessarily captured by visits conducted on a one-off basis. And they record the level of support to a degree that funding decisions hinge on: needing setup and standby while someone eats is a different level of support than needing physical assistance for the fine motor aspects of eating, and that is the difference between a vague plan and a defensible one. As they appear in reports, they are third-party observation over time, attributed and dated, sitting alongside clinical findings, and their concreteness in ordinary language makes them some of the most compelling paragraphs in the report.

How this plays out in practice

Imagine a man with MS who is funded based on determining the degree of change in his function. He presents well on assessment day, as he does at his best, and an assessment based on that visit would be an under-description of his needs. The designed version issued the observation frame to his support team three weeks before, and the data indicates a tremendous variation in function during the morning hours, physical help needed to transfer on approximately one-third of days, and two days where fatigue made function end by noon. In the report the assessment sits in the observed range, the fluctuation is not asserted but actually evidenced, and the plan that results funds his actual life, instead of just the good.

Next steps

Create your observation frame template this week, three tasks, three columns, one fortnight, and issue it prior to the next review of a supported participant. Then compare the report that you write with the one that you would have written. The practice change is to make that comparison.