The Algorithm for Mobility Assistive Devices: Can It Be Used in Documentation?
When the new power mobility device codes were initially introduced by CMS in 2006, a new National Coverage Determination provided us with a guide to help determine if an individual was appropriate for mobility assistive equipment (MAE), including power mobility. This series of yes/no questions that gave us a step-by-step thought process to rule out lower cost alternatives became known as the “algorithm”:
1. Does the beneficiary have a mobility limitation that significantly impairs his/her ability to participate in one or more MRADLs in the home
2. Are there other conditions that limit the beneficiary’s ability to participate in MRADLs at home? If so, can they be compensated sufficiently such that the provision of MAE will significantly improve the beneficiary’s ability to participate in MRADLs in the home?
3. Does the beneficiary or caregiver demonstrate the capability and willingness to consistently operate the MAE safely?
4. Can the functional mobility deficit be sufficiently resolved with a cane or walker?
5. Does the beneficiary’s typical environment support the use of wheelchairs including scooters/POVs?
6. Does the beneficiary have sufficient upper extremity function to propel an optimally configured manual wheelchair in the home to participate in MRADLs during a typical day?
7. Does the beneficiary have sufficient strength and postural stability to operate a POV/scooter?
8. Are the additional features provided by a power wheelchair needed to allow the beneficiary to participate in one or more MRADLs?
Manufacturers, clinicians and suppliers all began incorporating these questions into evaluation forms for physicians and therapists to use to document that the criteria for a power wheelchair had been met. However, CMS soon made it clear that merely answering these “yes/no” questions on a form was insufficient to verify eligibility for the MAE. Yet many evaluation forms continue to include the algorithm questions in either the objective section or the assessment section. In some cases, lines have been added for “comments” with the thought that this would provide adequate information to support the “yes/no” answers. So, can the algorithm be included in the physicians’ and therapists’ documentation and is it helpful? The answer to the first question is “yes, if you really want to”, while the answer to the second question is “it depends, but in most cases, no”.
First, let’s look at the true purpose of the algorithm. It is meant solely to provide a guide as to what the clinician should be looking at when evaluating for a mobility device. It basically walks someone through the process of starting with the simplest device (a cane) and systematically ruling out each device in the hierarchy until the appropriate one is identified. It is a thought process or an identification of the general “buckets” of areas that should be evaluated. But to sufficiently answer the questions, the clinician must perform specific physical and functional tests that result in objective, and preferably quantitative data. This is the only way that the clinician can arrive at an answer. And it is these objective findings that must be included in the documentation to substantiate the “yes/no” answers.
So what is the problem that we see when these questions are included in the documentation? As mentioned above, some clinicians are still merely answering “yes” or “no” without supportive comments. We know this is insufficient. But what if the documentation includes lines for “comments” to support the answer? Too often, the “comments” are too vague and too broad with no objective exam results to substantiate the conclusion. For example, Dr. X (or Therapist A) is recommending a power wheelchair for Mrs. S and answers “no” to the question “Can the functional mobility deficit be sufficiently resolved with a cane or walker?“. He then adds the comment “Mrs. S is no longer able to ambulate with a cane due to increasing lower extremity weakness and pain, abnormal gait and COPD”. However, there is nothing in this comment to quantify the weakness, indicate how far she can walk, indicate what respiratory effects she experiences after walking, quantify how long it takes for her to recover, or describe what her gait pattern looks like. It also does not provide similar information to describe her attempts to use a walker. So if this objective information is not found anywhere else in the documentation, this answer is is not sufficiently supported.
What should be included are comments similar to the following: “strength in the LEs is 3+/5 throughout on the right and 4-/5 throughout on the left; Pt is only able to ambulate 25’ using a cane before becoming significantly short of breath and it takes 3 minutes to recover; O2 saturation drops from 95% to 89% after walking 25’. She experienced similar respiratory effects after ambulating with a walker for 25’; Pt walks with a shuffling unsteady gait without an adequate heel strike which causes significant risk for falls; The pain in her shoulders due to her arthritis increased from 3/10 to 7/10 after using a walker for 25’.
Or let’s say the clinician answered “no” to the question “Does the beneficiary have sufficient upper extremity function to propel an optimally configured manual wheelchair in the home to participate in MRADLs during a typical day?” and added the comment “has tried propelling a manual wheelchair in her home but had difficulty due to UE weakness, fatigue, and pain from arthritis.” Again, this does not say how far she can propel, it does not quantify her weakness, pain or fatigue, it does not indicate specifically where the arthritis is (which joints), it does not indicate what type of manual wheelchair she attempted to propel and it does not indicate that the clinician even observed her attempting to propel it. If this information is not included in the objective exam results, then there is insufficient evidence to rule out manual mobility.
So the bottom line is that the algorithm questions are OK to use on an evaluation form or letter of medical necessity, however they are not sufficient by themselves to support the need for a mobility device, even with comments to support the “yes/no” answer, particularly if these comments are vague and general. There must be objective exam results, preferably quantitative if possible, to show that a sufficiently thorough physical and functional exam took place and to substantiate the “yes/no” answers. The specific physical limitations and capabilities of the individual must be identified and there must be information as to how these limitations and capabilities relate to the individual’s ability to perform daily activities.
On the other hand, we should not ignore these questions either. A reviewer will be looking for specific and sufficient information that does support the need for the recommended device and this includes what is outlined in the algorithm; that is, that the individual has physical and functional limitations and what they are; that the individual is unable to use a lower cost alternative and why; how the use of a lower cost alternative affects his/her ability to perform daily activities in a timely and safe manner; that the individual is able and willing to safely use the recommended device; and that the home is accessible for the device.
If you have provided a sample evaluation form to your referrals with the algorithm questions on it or if they are using one from someone else on a regular basis, make sure that they understand the purpose of the algorithm questions. Also make sure that they understand that the most important pieces of information are not necessarily the direct “yes/no” answers to these questions, but rather their (the therapists’ or physicians’) exam results and their assessment of these results.