The difference between contact guard assist (CGA) and standby assist (SBA) comes down to one thing: your hands. With contact guard assist, you keep a hand on the patient or on their gait belt, ready to steady or arrest a loss of balance, while the patient does 100% of the work. With standby assist, you take your hands off and stay close enough to help, but you provide no physical contact at all. Because contact guard involves touch and standby does not, a patient at the standby level is functionally more independent than one who still needs contact guard.
That is the short answer, and if it is all you needed, you can stop here. But the distinction gets slippery fast once you have to document it, defend it to a reviewer, or reconcile it with the way Medicare actually scores function today. This guide walks through both terms in plain language, places them on the full levels-of-assistance continuum, and shows how they map onto the two scoring systems occupational therapists live with: the classic FIM scale and the CMS Section GG scale that replaced it for Medicare. Everything here is tied to a primary source, because in a clinical note the difference between "supervision" and "min assist" is not just wording, it is a fall risk and a reimbursement claim.
Both contact guard assist and standby assist describe a patient who completes an activity under their own power. Neither one means you are lifting, supporting, or doing any of the physical effort. The only variable that separates them is whether you are touching the patient.
| Feature | Contact Guard Assist (CGA) | Standby Assist (SBA) |
|---|---|---|
| Physical contact | Yes: a hand on the patient or gait belt | No contact |
| Patient's effort | 100% | 100% |
| Your role | Steady or catch a loss of balance | Guard, cue, and be ready to step in |
| Also called | Contact guarding, "hands-on" guarding | Supervision, "standby" |
| Relative independence | Lower (still needs touch) | Higher (no touch needed) |
| FIM level | 4 (Minimal Contact Assistance) | 5 (Supervision or Setup) |
| Section GG code | 04 | 04 |
Notice the last two rows. Under the older FIM instrument the two levels are scored differently, a 4 versus a 5. Under Medicare's current Section GG coding scale, they collapse into the same code, 04, "Supervision or touching assistance." That single fact trips up more clinicians than any other part of this topic, and we unpack it in the FIM vs Section GG section below.
Contact guard assist, almost always abbreviated CGA (you will also see "CG" or "CGA x1" for one person guarding), means the therapist maintains light physical contact with the patient, usually a hand resting on a gait belt secured at the waist, while the patient performs the entire task themselves. The contact is protective, not assistive. You are not pulling them up, holding them, or bearing any of their weight. You are there so that if their knee buckles or they list to one side, your hand is already in position to steady them or to control a fall.
The gait belt is the tool of the trade here. The U.S. Department of Veterans Affairs Safe Patient Handling and Mobility program is explicit that a gait belt "aids/guides in ambulation of patients" and is not a lifting device. That is the whole philosophy of contact guard: the belt and your hand exist to guide and stabilize, not to haul. If you find yourself using the belt to lift or support the patient's weight, you have left contact guard and moved into hands-on assistance, which is a different level entirely.
Documentation shorthand: if you write "CGA," a reviewer reads "the patient did all the work; the therapist's hands were on the belt for safety only." If your note elsewhere describes you supporting the patient's trunk or lifting a limb, that contradicts CGA and invites a denial. Keep the picture consistent.
What does contact guard look like in practice? Picture a patient two days out from a total hip replacement taking their first lap down the hall with a front-wheeled walker. They can stand, step, and turn on their own, but their balance is not yet trustworthy. You walk just behind and to the side, one hand on the gait belt, the other hovering near their shoulder. You never actually assist a single step, but if they catch a toe, you are already holding on. That is textbook contact guard assist.
Standby assist, abbreviated SBA and frequently used interchangeably with "supervision," describes the same self-powered patient, minus the hand. You stay within arm's reach, you may give verbal cues or reminders ("watch that your walker is all the way forward before you step"), and you are ready to move in instantly, but you are not touching them. The patient has demonstrated enough balance and safety awareness that constant contact is no longer necessary, yet not enough that you would feel comfortable leaving the room.
The line between standby and true independence is safety and consistency. A patient who needs standby assist can usually complete the task correctly nine times out of ten, but that tenth attempt, when they forget a precaution, rush, or fatigue, is why you are still standing there. Once they can do it reliably and safely without you present, they graduate to supervision-free independence.
Standby assist and "supervision" are treated as the same level in most rehab settings, but some facilities reserve "supervision" for cueing at a slight distance and "standby" for being right at the patient's elbow. When in doubt, follow your facility's operational definitions and document the specific reason you are there (cueing, balance, safety awareness) rather than relying on the label alone.
Contact guard and standby are two rungs on a longer ladder that runs from complete independence down to total dependence. Occupational therapists use this continuum to describe function in daily-living tasks, transfers, and mobility, and it is the backbone of almost every rehab progress note. Here is the full continuum from most to least independent, with the everyday term, what the patient contributes, and how each rung lines up with the two scoring systems covered later.
| Level (Most to Least Independent) | What the Patient Does | Therapist's Role | FIM Level | Section GG Code |
|---|---|---|---|---|
| Independent (I) | 100%, safely, no device needed | None | 7 | 06 |
| Modified Independent (Mod I) | 100%, but needs a device, extra time, or safety consideration | None, but a device or adaptation is used | 6 | 06 |
| Supervision / Standby (SBA) | 100%, with cueing or someone nearby | Watch, cue, no contact | 5 | 04 |
| Contact Guard (CGA) | 100%, but needs hands-on-belt for balance | Touch or steady, no effort | 4 | 04 |
| Minimal Assist (Min A) | 75% or more of the effort | Up to 25% of the effort | 4 | 03 |
| Moderate Assist (Mod A) | 50% to 74% of the effort | 26% to 50% of the effort | 3 | 03 |
| Maximal Assist (Max A) | 25% to 49% of the effort | More than half the effort | 2 | 02 |
| Dependent / Total Assist (Dep) | Less than 25%, or none | Does all or nearly all the work; may need two people | 1 | 01 |
A few things are worth pausing on. First, contact guard and minimal assist share the same FIM level (4) but split apart under Section GG: pure contact guard is a 04, while minimal assist, where your hands actually do a little of the work, drops to a 03. Second, there is no official CMS crosswalk that translates FIM levels into Section GG codes, so treat the last two columns as an approximate conceptual alignment, not a billing formula. The two systems were built for different purposes, which is exactly why they do not line up cleanly.
Once a patient needs you to physically do part of the task, the levels are defined by how much of the effort each of you contributes. These are the numbers every new clinician memorizes:
These percentage bands come straight from the FIM instrument's scoring criteria, and they remain the common language of rehab documentation even in settings that no longer score the FIM formally. It is worth being honest about where they originate: they are FIM-derived thresholds, not a separate CMS or AOTA standard. You will occasionally see the moderate band written as "50% to 75%" instead of "50% to 74%," a small rounding difference between published tables: the AAPM&R functional assessment table writes it as 50% to 75%, while the canonical FIM manuals use 50% to 74%. Pick one and stay consistent within your documentation.
The percentages are an estimate of physical effort, not a stopwatch measurement. "50%" is your clinical judgment of how much of the work the patient did, and reasonable clinicians disagree. That subjectivity is a known limitation of these scales, which is why your narrative description matters more than the number.
Here is the part almost every competing article skips, and the part that most affects your daily notes. There are two different systems for scoring function, and contact guard versus standby lands differently in each.
The Functional Independence Measure (FIM) is a seven-level scale developed and owned by the Uniform Data System for Medical Rehabilitation (UDSMR). From 7 (complete independence) down to 1 (total assistance), it is the framework that gave us the min/mod/max vocabulary. Critically for this article, the FIM draws a hard line between our two terms: standby and supervision are level 5, while contact guard is named explicitly at level 4, "minimal contact assistance," defined as the patient needing "no more than touching" and expending 75% or more of the effort. So on a FIM-scored chart, moving a patient from CGA to SBA is a documented, meaningful jump from a 4 to a 5.
The catch: Medicare retired the FIM. For inpatient rehabilitation facility discharges on or after October 1, 2019, the FY2019 IRF PPS Final Rule removed the FIM instrument and its function modifiers from the IRF Patient Assessment Instrument. The FIM still exists as a proprietary, UDSMR-licensed instrument and many facilities use it internally, but it is no longer the federal reporting standard.
What replaced it is Section GG, a six-level scale for self-care and mobility. Under the IMPACT Act of 2014, CMS standardized functional-status data across post-acute settings, so the same Section GG items now appear on the IRF-PAI (rehab hospitals), the MDS (skilled nursing), the OASIS (home health), and the LTCH CARE Data Set. The scale runs:
When an activity cannot be assessed, Section GG uses "activity not attempted" codes instead: 07 (patient refused), 09 (not applicable), 10 (not attempted due to environmental limitations), and 88 (not attempted due to medical condition or safety concerns).
Now the key insight. Read the official definition of code 04 from the CMS coding guidance:
04. Supervision or touching assistance: Helper provides verbal cues and/or touching/steadying and/or contact guard assistance as patient/resident completes activity. Assistance may be provided throughout the activity or intermittently.
Verbal cues (that is your supervision and standby), touching or steadying, and contact guard assistance are all bundled into the one code. Under Section GG, contact guard assist and standby assist are the same score: 04. The AOTA Section GG assessment form reflects the same collapsed definition. So the CGA-versus-SBA distinction that mattered so much on the FIM has no effect on the Section GG code your facility reports to Medicare.
This does not make the distinction pointless. Section GG is a snapshot for federal quality reporting; your treatment note is a clinical and legal record of the patient's real status and trajectory. Documenting the move from contact guard to standby still shows progress, justifies skilled care, and guides the next clinician, even when the reported code does not budge.
One more nuance that catches people: Section GG code 03 is where physical assistance actually begins, and the split between 03 and 02 is not a percentage, it is simply whether the helper does less than half or more than half of the effort. If you carry the FIM percentage bands into Section GG and write "moderate assist, patient did 60%," you are mixing two systems. Section GG does not use percentages at all.
Grading a patient correctly is a clinical judgment, and it is one of the more consequential calls you make, because it drives the plan of care and, downstream, the safety of the next person who works with them. A few questions help you land on the right level:
This last point is where occupational therapy's grading-the-activity mindset shines. Progressing a patient from moderate assist to min assist to contact guard to standby to independent, one graded step at a time, is the visible arc of recovery that your documentation should tell.
Safety reminder grounded in the evidence: if a patient does begin to fall from standing, guidance from safe-patient-handling programs is to control and lower them to the floor rather than trying to catch or hold them upright, which is a leading cause of caregiver back injury. Your body is not a safe braking system, which is exactly why contact guard uses a gait belt and good positioning rather than brute strength.
The assist level is one of the most scrutinized elements of a therapy note, because it is the shorthand payers use to judge whether skilled intervention was necessary and whether the patient is progressing. Sound documentation pairs the label with a specific, observable picture. Compare these two lines for the same session:
The strong version tells a reviewer why skilled care was needed, what specifically the patient could and could not do, and where they sit on the continuum. If you want a deeper treatment of note structure, defensibility, and payer expectations, see our full guide to occupational therapy documentation, and if the abbreviations here are new, our reference on occupational therapy medical abbreviations spells out CGA, SBA, Min A, and the rest.
Two documentation habits protect you. First, describe the reason for the assist (balance, weakness, cognition, safety awareness), not just the amount. Second, keep the assist level consistent across your note: a patient who is "CGA for ambulation" should not also be described as needing you to support their trunk, because that is hands-on assistance, and the internal contradiction is exactly what a reviewer flags.
A handful of errors show up again and again when clinicians grade or document these levels:
Contact guard assist means you keep a hand on the patient or their gait belt for balance while they do all the work; standby assist means you stay close and may cue them but provide no physical contact. Because standby involves no touch, it represents a higher level of independence than contact guard.
CGA stands for contact guard assist (sometimes written CG or CGA x1 for one person). It is a level of assistance in which the therapist maintains light protective contact, typically on a gait belt, while the patient performs 100% of the activity. Under Medicare's Section GG scale it is coded 04, "supervision or touching assistance."
Imagine a post-surgical patient walking the hall with a walker while the therapist walks alongside with one hand on the gait belt, never assisting a step but ready to steady them if they lose balance. The patient is doing everything; the hand is there only for safety.
Contact guard is used precisely because the patient carries a fall risk, so the main risk is a loss of balance or fall during the activity. Guarding, a gait belt, and correct positioning let you control or lower a fall rather than let the patient hit the floor. There is also a risk of over-guarding, which can make a patient dependent on your presence and slow their progress toward standby and independence.
In most rehab settings the two are used interchangeably to mean the patient completes the task without physical contact while someone stays nearby to cue or intervene. Some facilities distinguish "supervision" (cueing from a short distance) from "standby" (right at the patient's side), so follow your setting's operational definitions.
No. On the current Section GG scale, verbal cues, touching or steadying, and contact guard assistance all map to the same code, 04. The distinction between them matters clinically and for the FIM, but it does not change the Section GG code reported to Medicare.