Falls With Major Injury Sit at 0.93%. The Warning Signs Arrive by Phone First.

Falls With Major Injury Sit at 0.93%. The Warning Signs Arrive by Phone First.

CMS's July 2026 national home health file puts the rate of falls with major injury at 0.93% of home health quality episodes. That is the number on Care Compare. It is accurate, it is the number surveyors and referral partners see, and it is close to the floor of what any quality measure can usefully discriminate.

It is also the wrong number to manage against, because by the time an episode counts toward it, the fracture has already happened. The signal that a patient was heading there almost always arrived earlier, and in most agencies it arrived on the phone.

What 0.93% actually counts

The measure is the Application of Percent of Residents Experiencing One or More Falls with Major Injury, and its specification is narrower than the name suggests. Per the CMS Home Health QRP measure calculations manual, the numerator is home health quality episodes in which the patient experienced one or more falls since start or resumption of care that resulted in major injury, captured through OASIS items J1800 and J1900C. Major injury is defined specifically: bone fractures, joint dislocations, closed-head injuries with altered consciousness, or subdural hematoma.

Three things follow from that definition, and operations leaders should be clear on all three.

The measure captures	The measure does not capture
Falls with fracture, dislocation, closed-head injury with altered consciousness, or subdural hematoma	Falls with bruising, lacerations, sprains, or no injury at all
Falls occurring during the home health episode, from SOC/ROC forward	Falls before admission or after end of care
Falls documented on the OASIS assessment	Falls the patient never mentioned to a clinician

That last row is the operational one. The measure is only as complete as what patients disclose and what staff record. A fall that a daughter mentions to your intake coordinator on a Tuesday afternoon, that never makes it into a clinical note, does not exist as far as the measure is concerned — and, more to the point, does not exist as far as your care plan is concerned either.

The falls that never reach the measure

The population-level picture from CDC is far larger than what home health public reporting captures. According to CDC's older adult falls data:

• More than one out of four older people falls each year.
• Less than half tell their doctor.
• One out of ten falls results in an injury, and about 37% of those who fall reported an injury requiring medical treatment.
• Falls drive about 3 million emergency department visits and about 1 million hospitalizations a year.
• Nearly 319,000 older people are hospitalized for hip fractures annually.
• Falling once doubles the chance of falling again.

Put the CDC figure next to the CMS figure and the gap is the whole problem. More than a quarter of your census is falling. Under one percent of episodes produce a documented major-injury fall. Almost everything in between is invisible to your quality data — and roughly half of it is invisible to the clinician too, because the patient never brought it up in a clinical setting.

But patients and families do bring it up. Just not to the nurse, and not in a documented visit. They bring it up on the phone, in the middle of a call about something else.

What a pre-fall call sounds like

The following are illustrative examples, not customer data. They describe the shape of calls that experienced intake and scheduling staff recognize immediately and that call logs almost never preserve.

The reschedule with a reason attached

A daughter calls to move Thursday's visit to Friday. Somewhere in the middle she says her mother "had a rough night, she slipped getting to the bathroom but she's fine." The call is dispositioned Schedule Change. The reschedule is honored. The slip is gone.

The medication question that is really a dizziness report

A patient calls asking whether she should keep taking a new blood pressure medication because she has been "light-headed in the mornings." The coordinator correctly routes her to the nurse for a callback. The callback happens two days later. In between, nobody has connected the light-headedness to the fact that this patient also has a walker she has stopped using.

The environment complaint

A family member calls to complain that the aide left the hallway rug bunched up, or that the shower chair is wobbly, or that the porch light has been out for two weeks. These get logged as service complaints. Each one is also a fall-risk finding about the physical environment, and none of them is routed anywhere a clinician will read it.

None of these three calls is a fall report. All three are fall risk, arriving in the format the agency actually receives most of its information in. We have written before about the specific fall-related call patterns worth monitoring; the underlying problem is that the phone is the highest-volume clinical data stream in most agencies and the least structured.

What to do this week

You do not need new software to make progress on this. You need to change what gets captured and where it goes.

1. Add one required field to your call intake form. A single yes/no: "Did the caller mention a fall, near-fall, dizziness, or unsteadiness?" Not a free-text box. A field that can be counted and filtered. If your system supports it, make it mandatory before a call can be closed.
2. Define what happens when it is checked. The failure mode is capturing the flag and doing nothing with it. Write down the rule: flagged call goes to the case manager's queue same day, clinician contacts patient within 24 hours, outcome documented. If the answer is "we would call them anyway," write that down too, then check next month whether it happened.
3. Pull your last 30 days of calls dispositioned as reschedules, complaints, and medication questions. Read 50 of them. Not to audit anyone — to find out how often fall language is sitting inside a call that was closed as something else. Most agencies that do this exercise are surprised by their own numbers.
4. Cross-check against your OASIS J1800 responses. If your call records show fall mentions for patients whose end-of-care assessment shows no falls, you have a documentation gap that is also a care gap. That reconciliation is a twenty-minute exercise on a sample.
5. Brief the aide team separately. Aides see the home more often than anyone. Give them the same one-question prompt and a low-friction way to report it that is not a phone call to a busy scheduler.

Steps one through three are achievable in a single week by one person with access to your telephony reports and your EMR. Step four is the one that tells you whether the first three matter.

Why this is a measurement problem before it is a clinical one

Home health agencies are not short on fall prevention protocols. They are short on the inputs those protocols depend on. A fall risk assessment completed at start of care describes a patient on one day; the patient's actual risk changes with a new prescription, a hospital stay, a caregiver who stopped coming, or a rug. Those changes get reported by phone, between visits, to whoever picks up.

The gap between your Care Compare number and the CDC's population number is not a scoring artifact. It is the volume of information your agency already receives and does not retain in a form anyone can act on. Reporting-level measures like the 0.93% national rate tell you what happened across an industry last year. They were never built to tell you which patient on your census is trending the wrong way this week — that is a different kind of instrument, closer to what we have described as the difference between dashboards and surfaces.

SurfacerIQ works in this category: interaction intelligence that reviews call content across the full volume of patient and family conversations rather than a sampled few, and surfaces clinical and compliance signals that would otherwise stay inside a disposition code. That is the capability class; whether you build it, buy it, or do it with a spreadsheet and a discipline, the requirement is the same.

This post is general information about publicly reported quality measures and is not legal or clinical advice. Measure specifications change; verify against the current CMS QRP manual before making documentation decisions.

If you want to talk through what your call data would show, talk to us.