| Research TBI Model Systems |
| An Analysis of FIM Score Rating Difficulties that can Affect Rehabilitation Payments |
| The Question: Does clinician expertise and the degree of difficulty for rating a Functional Independence Measure (FIM) item explain the clinician’s overconfidence in their FIM rating accuracy? Past Studies Medicare and Medicaid use a Prospective Payment System (PPS) that is intended to transfer the financial risk to the healthcare provider, instead of the payer. This is accomplished by paying healthcare providers upfront by case complexity rather than paying them at patient discharge based on the patient’s length of stay. By using a PPS, clinicians have an incentive to work efficiently. Inpatient rehabilitation services have their own form of PPS called Case-mix groups (CMGs). Case-mix groups are designed to forecast the cost of care for each individual that enters inpatient rehabilitation. CMGs are figured by using an individual’s functional status at admission to rehabilitation. 𠇏unctional status” refers to how independently an individual can perform activities needed for daily living. “Activities of daily living” are the very basic self-care activities that individuals perform to take care of themselves during a typical day. Examples include toileting, bathing, eating, and dressing; thinking skills, such as the ability to remember things or figure things out; moving the body from place to place, such as walking or getting in and out of bed; and communication skills. Individuals with traumatic brain injury can experience difficulty performing activities of daily living because brain injuries can affect the way they think, act, feel, and move their bodies. Clinicians use an assessment called the Functional Independent Measure (FIM) to rate, record, and measure an individual’s ability to perform activities of daily living. The FIM score is based on the clinician’s observation of the individual performing the tasks. Because FIM scores are based on the clinician’s observation and interpretation, it is not a very specific process and there may be variance in the way clinicians score the same individual. It appears that there is a general tendency for health care providers to be overconfident in their ratings, with those with less expertise having a greater tendency to be so. It also appears that when the clinicians rate items on the FIM that are more difficult to judge, their levels of overconfidence increases. Past studies show that clinicians that are overconfident in their abilities to rate FIM scores or that do not use the FIM tools designed to assist with rating decisions are more likely to make inaccurate ratings. Therefore, the accuracy of a CMG based PPS in determining payment to a facility is limited by the accuracy of the clinicians that provide the FIM scores. It has been suggested that such a payment system that relies on the healthcare providers’ behavior and opinions may tend to favor over paying the facility and jeopardize the cost-effectiveness of the CMG PPS as a whole. Research is lacking regarding what influences clinicians’ overconfidence and their ability to make accurate or inaccurate FIM score ratings. This Study included 50 clinicians, including physical therapists, occupational therapists, rehabilitation psychologists, speech pathologists and rehabilitation nurses from three inpatient rehabilitation centers. The participants answered 60 true/false FIM questions. For each question, they completed a scale to test the confidence in the accuracy of their responses. The participants also indicated at which point they would consult a colleague, manual, or flow chart to help them make a decision. This study included expert and novice clinicians, as well as hard and easy questions. The researchers statistically analyzed the results to determine the relationship between clinician accuracy and confidence. Overall, the clinicians were overconfident in the accuracy of their responses. The clinicians with less expertise were more likely to be overconfident. Clinicians showed less overconfidence for items that were easier to rate on the FIM than for items that were more difficult. The clinicians indicated that they would be most likely to consult a colleague, manual, or flow chart when their confidence for a response was 78.9%. When the clinicians were 78.9% confidence in their responses, their accuracy rating was approximately 65%. When the clinicians reported that they had 100% confidence in their responses, the responses were actually 78.9% correct. The researchers found that items on the FIM pertaining to thinking skills were the most difficult for the clinicians to rate accurately, and that those items involving body movement skills were the easiest. The clinicians also had more difficulty rating midrange FIM scores accurately than they did the scale’s end range scores. Who May Be Affected By These Findings: Clinicians, payment specialists, and individuals with traumatic brain injuries Caveats: Although the results of this study are similar to former studies, the researchers caution that the results may not be generalizable because they relied on the clinicians’ self-reports instead of observation. Bottom Line: Overall, the clinicians were overconfident in the accuracy of their responses. The clinicians with less expertise were more likely to be overconfident. The clinicians showed less overconfidence for items that were easier to rate on the FIM than for items that were more difficult. The researchers state that these clinicians should receive training to reduce overconfidence or that they should always double-check their ratings with the FIM decision making tools. Improved accuracy of FIM scoring would improve the quality of CMG ratings and increase correct payment for inpatient rehabilitation services within the PPS. Find This Study: Doctor, J.N., Wolfson, A.M., McKnight, P., & Burns, S.P. (2003). The effect of inaccurate FIM instrument ratings on prospective payments: A study of clinician expertise and FIM rating difficulty as contributing to inaccuracy. Archives of Physical Medicine, 84, 46-50. |
Please take a moment to comment on the value of this abstract: |