Preventing Falls in Hospitals

A basic principle of quality measurement is: If you can't measure it, you can't improve it. Therefore, fall rates and fall prevention practices must be counted and tracked as one component of a quality improvement program. By tracking performance, you will know whether care is improving, staying the same, or worsening in response to efforts to change practice. Moreover, continued monitoring will help you understand where you are starting from and whether your improvement gains are being sustained.

During the course of your fall prevention improvement effort and on an ongoing basis, you should regularly assess your fall rates and fall prevention practices. We recommend that you regularly monitor: (1) an outcome (such as falls per 1,000 occupied bed days), (2) at least one or two care processes (e.g., assessment of fall risk factors and actions taken to reduce fall risk), and (3) key aspects of the infrastructure to support best practices (e.g., checking for interdisciplinary participation in Implementation Team).

The questions below will help you and your organization develop measures to track fall rates and fall prevention practices:

• How do you measure fall and fall-related injury rates?
• How do you measure fall prevention practices?

Measuring Fall Rates and Fall Prevention Practices: Locally Relevant Considerations Your hospitals may experience challenges in trying to measure fall rates and fall prevention practices, such as: Revising incident reports to include more specific fields for contributing factors to falls (e.g., high-risk medications, which medications). Finding mechanisms to communicate fall incident report information to the Implementation Team. Using process metrics to measure the adherence to fall prevention strategies. Spreading lessons learned from postfall safety huddles and root cause analyses from one hospital unit to another. This section will discuss these types of issues.

Return to Contents

5.1. How do you measure fall and fall-related injury rates?

5.1.1. Why measure fall and fall-related injury rates?

Fall and fall-related injury rates are the most direct measure of how well you are succeeding in making patients safer related to falls. If your rates are improving, then you are likely doing a good job in preventing falls and fall-related injuries. Conversely, if your fall and fall-related injury rates are getting worse, then there might be areas in which care can be improved. You can use these data to make a case for initiating a quality improvement effort and monitoring progress to sustain your improvements.

5.1.2. What should be counted?

In measuring fall rates, you will need to count the number of falls and the number of occupied bed days on your unit over a given period of time, such as 1 month or 3 months. To count falls properly, people in your hospital or hospital unit need to agree on what counts as a "fall." Defining a fall is especially a problem in "borderline" cases, such as when a patient feels her knees giving out while walking with a hospital staff member and the staff member eases the patient onto the floor. Also, staff may feel pressure to underreport borderline cases because of concern that their unit will compare poorly with other units. Therefore, when a uniform definition of fall is shared throughout the hospital, it needs to be coupled with a culture of trust in which reporting falls is encouraged. There are many definitions of falls, and you should choose one appropriate for your situation.

You may also want to track the number of repeat falls on your unit. Sometimes a single repeat faller can skew the fall rate for the entire unit, so knowing about repeat falls can be helpful in understanding your data.

With each fall, you will need to define the level of injury that occurred, if any. Combining information about falls with the level of injury can give you an injurious fall rate. The injurious fall rate can be tracked just like the total fall rate. The advantage of the injurious fall rate is that it tracks the more clinically important falls and is less likely to be affected by the "borderline" falls problem noted above. The disadvantage is that if there are relatively few injurious falls compared with total falls, it will be hard to tell whether your fall prevention program is making a difference with respect to injuries. Thus, we recommend that both total and injurious fall rates be computed and tracked.

The National Database of Nursing Quality Indicators (NDNQI) Data Web site (https://www.nursingquality.org/data.aspx ) has a link in the bottom right corner titled "ANA is the NQF measure steward." This link takes you to definitions of falls and patient days so that fall rates may be calculated. A paraphrase of the March 2012 NDNQI fall definition follows: A patient fall is an unplanned descent to the floor with or without injury to the patient. Include falls when a patient lands on a surface where you wouldn't expect to find a patient. All unassisted and assisted falls are to be included whether they result from physiological reasons (fainting) or environmental reasons (slippery floor). Also report patients that roll off a low bed onto a mat as a fall. A synopsis of the NDNQI definition for repeat fall follows: More than one fall in a given month by the same patient after admission to this unit, may be classified as a repeat fall. The NDNQI definitions for injury follow: None—patient had no injuries (no signs or symptoms) resulting from the fall, if an x-ray, CT scan or other post fall evaluation results in a finding of no injury. "Minor—resulted in application of a dressing, ice, cleaning of a wound, limb elevation, topical medication, bruise or abrasion. Moderate—resulted in suturing, application of steri-strips/skin glue, splinting or muscle/joint strain. Major—resulted in surgery, casting, traction, required consultation for neurological (basilar skull fracture, small subdural hematoma) or internal injury (rib fracture, small liver laceration) or patients with coagulopathy who receive blood products as a result of the fall. Death—the patient died as a result of injuries sustained from the fall (not from physiologic events causing the fall)."

Determine whether staff know the definition of falls and injuries that your hospital has selected.

5.1.3. What measures do you use for fall rates?

The best measure of falls is one that can be compared over time within a hospital unit to see if care is improving. Sometimes staff would like to simply track the number of falls that occur every month or every quarter on a given unit. The problem with only tracking falls is that this does not account for how full or empty the unit was at any given time. If the unit census is running low, there will be fewer falls, regardless of the care provided. Therefore, we recommend that you calculate falls as a rate, specifically, the rate of falls per 1,000 occupied bed days. Later, we will show you how to make this calculation. You can similarly calculate the rate of injurious falls per 1,000 occupied bed days.

There is no single "right" approach to measuring fall rates. Every approach has advantages and disadvantages. While we make specific recommendations below, the most important point is to be consistent. Rates calculated by one approach cannot be compared with rates calculated another way.

Assess whether unit staff understand the difference between number of falls versus a fall rate. Define the measurement approach that you will use, and use it consistently throughout the hospital.

5.1.4. What do you need to calculate fall and fall-related injury rates?

To calculate fall and fall-related injury rates, whether at the unit level or at the overall facility, you need to know who fell, when the fall occurred, and what the degree of injury was, if any. You also need to know the daily census on the unit where you would like to calculate the fall rate, or throughout the hospital if you are calculating a fall rate at the hospital level. To obtain this information, you must complete two tasks:

1. Generate an incident report for every fall that occurs. The incident report will need to contain, at a minimum:
   • The fact that the incident being reported was a fall.
   • The patient in whom the fall occurred.
   • The date the fall occurred.
   • The unit the patient was assigned to at the time of the fall.
   • The location of the fall.
   • A detailed report about the circumstances of the fall.
   • The level of injury, if any.
   The circumstances of the fall and level of injury will be important as well for analysis, discussed later. But for calculation of a fall rate, you need the date the fall occurred and the responsible unit (if you want to calculate a unit fall rate).
2. Determine whether your hospital information system can provide you with the average daily census on the unit of interest, or in the hospital, for the time period over which you want to calculate a fall rate. The average daily census is the number of beds, on average, that are occupied throughout the day. Because patients come and go quickly on many hospital units, if you have access to a computerized system to give you the daily census, this will simplify your life later. If not, you will need to choose a point in time each day that is convenient to check the number of occupied beds on your unit, and write down that number each day, to be tallied as explained below.

Learn more about your hospital's incident reporting system. Some hospitals have electronic incident reporting systems that will make it easier to count the number of falls that have occurred on your unit or in your hospital.

To learn how the National Database of Nursing Quality Indicators (NDNQI) recommends capturing data on falls and patient-days, refer to the link titled "ANA is the NQF measure steward" at the NDNQI Data Web site: (https://www.nursingquality.org/data.aspx ). This will take you to the document Guidelines for Data Collection on the American Nurses Association's National Quality Forum Endorsed Measures. If information technology personnel are developing an electronic incident reporting system, they may find the Pennsylvania Patient Safety Authority's standard structure for incident reporting useful: See section 2.8 (page 60) of http://patientsafetyauthority.org/PA-PSRS/Documents/part2-xmldocumentdefinition.pdf [Plugin Software Help] . The AHRQ Common Formats Web site also links to a standard structure for collecting data for a fall-related incident report: https://www.psoppc.org/web/patientsafety/version-1.2_documents#Fall .

5.1.5. How do you calculate fall rates?

We recommend fall rates be calculated monthly based on the information from incident reports and daily census discussed above, but quarterly may also be appropriate. The advantage of monthly data over quarterly is that you have more regular opportunities to feed data back to staff about their improvements. The disadvantage is that it requires more effort to review data monthly rather than quarterly.

Let's say, as an example, that you want to calculate the fall rate for the month of April on a 30-bed unit. Rates are calculated as follows:

• First, count the number of falls that occurred during the month of April from your incident reporting system. Let's say there were three falls during the month of April.
• Then figure out, for each day of the month at the same point in time, how many beds were occupied on the unit. For example, on April 1, there may have been 26 beds occupied; on April 2, there may have been 28 beds occupied, and so on. The hospital may have a way of reporting this information to you (for example, midnight census).
• Add up the total occupied beds each day, starting from April 1 through April 30. Let's say the total adds to 879 (out of a maximum of 900, since if all 30 beds were occupied on all 30 days, 30 x 30 would equal 900). If your hospital can calculate for you the total number of occupied bed days experienced on your unit during the month of April, then you can just use this number, skipping step number 2.
• Divide the number of falls by the number of occupied bed days for the month of April, which is 3/879= 0.0034.
• Multiply the result you get in #4 by 1,000. So, 0.0034 x 1,000 = 3.4. Thus, your fall rate was 3.4 falls per 1,000 occupied bed days.

Identify a person or team in the organization who will be responsible for these calculations. Identify the sources of data that this person or team will use. If current data are not available or are not accurate, develop a strategy for improving data quality.

5.1.6. How should you use the monthly data on fall rates?

Use the information on fall rates that you collect in three ways.

First, examine your rates every month and look at the trend over time. How are they changing? Are they improving or getting worse? Can you relate changes in your fall rate to changes in practice? Think about what you have or have not been doing well over the past months and relate it to whether the fall rate is getting better or worse.

Remember that fall rates may change based on the season of the year and can be quite different from unit to unit (e.g., geriatric psychiatry unit versus intensive care unit). Don't overreact to any individual month's data as there can be fluctuations from month to month. Focus on the underlying trend of the data over time and whether fall rates are increasing or decreasing.

Graphing your data in a run chart is a good way to visually examine trends in the fall rate. A run chart looks like this:

In this case, the fall rate is plotted on the vertical axis and the month of the year is plotted from left to right.

A run chart like the one above can be created using a template available at no cost after free registration at the Institute for Healthcare Improvement Web site: www.ihi.org/knowledge/Pages/Tools/RunChart.aspx . The template is a downloadable, easy-to-use spreadsheet that allows you to enter your data. The spreadsheet also includes a tab with tips for interpreting your run chart.

When you first implement a quality improvement program and begin tracking performance, increased fall rates are frequently seen. This is not necessarily related to worse care. Instead, unit staff members are becoming better at reporting falls that were previously missed. This is another reason it is equally important to track fall-related injuries at the same time.

One study, using data from the National Database of Nursing Quality Indicators, found that fall rates varied substantially across units: Intensive Care Unit: 1.30 falls/1,000 patient days. Surgical: 2.79 falls/1,000 patient days. Stepdown: 3.44 falls/1,000 patient days. Medical-Surgical: 3.92 falls/1,000 patient days. Medical: 4.54 falls/1,000 patient days. Rehabilitation: 7.15 falls/1,000 patient days. For more information, see Lake ET, Shang J, Klaus S, et al. Patient falls: association with hospital magnet status and nursing unit staffing. Res Nurs Health 2010;33:413-25.

Further reading for those who want a more indepth look at how to collect and analyze data on fall rates: For a general overview of how to collect and use data for quality improvement: Needham DM, Sinopoli DJ, Dinglas VD, et al. Improving data quality control in quality improvement projects. Int J Qual Health Care 2009;21(2):145-50. To learn how to create a basic control chart for falls, see section titled "The u-chart" in Mohammed MA, Worthington P, Woodall WH. Plotting basic control charts: tutorial notes for health care practitioners. Qual Saf Health Care 2008;17:137-45. To analyze data on rare events, such as injurious falls, learn about the g-type control chart in Benneyan JC. Number-between g-type statistical quality control charts for monitoring adverse events. Health Care Manage Sci 2001;4:305-18. For an overview of how to calculate rates, identify trends, and present data: Quigley P, Neily J, Watson M, et al. Measuring fall program outcomes. Online J Issues Nurs 2007;12(2). Available at: www.nursingworld.org/MainMenuCategories/ ANAMarketplace/ANAPeriodicals/OJIN/ TableofContents/Volume122007/No2May07/ArticlePreviousTopic/ MeasuringFallProgramOutcomes.aspx .

The second way to use your data on falls is to disseminate the information to key stakeholders and to unit staff. Post monthly rates in places where all staff can see how the unit is doing. Send reports to leadership. Dissemination of information on performance is critical to your quality improvement effort.

The third way to use your data is to study in detail what led to the occurrence of each fall, particularly falls resulting in injury. Try to understand why the fall occurred and how such an incident might be prevented in the future. In particular, try to determine whether the falls are irregular events (e.g., a patient's first-ever seizure that resulted in a fall) or whether there is a regularity to the types of falls (e.g., related to toileting) that suggest a specific intervention is needed to improve care.

To get an idea of how incident report data can be used to better understand the circumstances of falls in a hospital, see this article: Hitcho EB, Krauss MJ, Birge S, et al. Characteristics and circumstances of falls in a hospital setting: a prospective analysis. J Gen Intern Med 2004;19:732-9.

Root cause analysis is a useful technique for understanding reasons for a failure in the system. Root cause analysis is a systematic process during which all factors contributing to an adverse event are studied and ways to improve care are identified. If you are not familiar with root cause analysis, work with your quality improvement department to learn how to conduct this analysis.

There are two different kinds of root cause analyses: aggregate and individual. For an aggregate analysis, the Implementation Team would review all falls, or all falls with injury, that occurred over the previous month, quarter, or year, for example. Using incident report information that is collected in a standard fashion, the team would seek to determine the main causes of falls in the hospital or on specific units, and then implement changes to address these causes. Often someone within the hospital's Quality Management (or similar) department can help in creating reports that can be reviewed as part of an aggregate root cause analysis.

An individual-level root cause analysis can occur after any fall, particularly falls with injury. Individual-level root cause analyses are carried out by the Unit Team immediately after a fall. These analyses can take the form of a postfall safety huddle, which is an informal gathering of unit staff to discuss what caused the fall and how subsequent falls or injuries can be prevented (go to section 3.4.4 for details).

	Sample postfall huddle forms may be found at the Minnesota Hospital Association Web site: www.mnhospitals.org/Portals/0/Documents/ptsafety/falls/post-fall-huddle-revised.pdf [Plugin Software Help] www.mnhospitals.org/Portals/0/Documents/ptsafety/falls/post-fall-huddle-documentation.pdf [Plugin Software Help]   You can use these forms or create your own, based on your hospital's specific needs. You can also build a form based on the postfall assessment form for root cause analysis (Tool 3O) in this toolkit. The key is to do a thorough assessment, identify the causes contributing to the fall, and come to a decision about actions that need to be taken to prevent a fall or injury in the future. Data should be collected in a standardized fashion, which should include all the data needed to complete an incident report. Standard data structures for incident reports may be found in the resource box in section 5.1.4.
	A primer on root cause analysis is available on the AHRQ Patient Safety Network Web site at: http://psnet.ahrq.gov/primer.aspx?primerID=10. For additional information and tools about root cause analysis, see the Veterans Affairs National Center for Patient Safety Web site at: www.patientsafety.gov/vision.html#rca. For tools, go to: www.patientsafety.gov/CogAids/RCA/index.html#page=page-1. For a step-by-step guide to aggregate root cause analysis: see Neily J, Ogrinc G, Mills P, et al. Using aggregate root cause analysis to improve patient safety. Jt Comm J Qual Patient Saf 2003;29(8):434-9.
	Identify audiences for the data at different levels of the organization and determine through which paths you will provide the data. For example, for senior managers, report the data in a leadership meeting or performance improvement committee meeting. Assess whether unit staff know the unit's fall and fall-related injury rate and whether it is improving over time.

5.1.7. Are there national benchmarks you can use for comparison with your fall rates?

The question of how well your hospital is performing relative to other hospitals often arises. Unfortunately, there are no national benchmarks with which you can compare your performance. In part this is due to the difficulties in making sure patients are similar across hospitals, since some patients are more likely to fall than others and hospitals care for different types of patients. Therefore, we encourage you to focus more on improvement over time within your units and your hospital overall, rather than focusing strictly on your hospital's performance compared with an external benchmark.

That having been said, there are a number of ongoing initiatives to determine fall rates using a standardized method across a large number of hospitals. These include the National Database of Nursing Quality Indicators, the Collaborative Alliance for Nursing Outcomes, and the Centers for Medicare & Medicaid Services (CMS) reporting on falls with trauma occurring in hospitals.

Learn more about ongoing data collection initiatives: National Database of Nursing Quality Indicators, Data: https://www.nursingquality.org/data.aspx . Collaborative Alliance for Nursing Outcomes, CALNOC Registry: http://calnoc.org/displaycommon.cfm?an=1&subarticlenbr=8 . For more information on the public reporting of falls with trauma that occur in hospitals participating in the Medicare program, go to the U.S. Department of Health and Human Services Hospital Compare Web site (www.hospitalcompare.hhs.gov/Data/RCD/Hospital-Acquired-Conditions.aspx) and look for the link to the Hospital Acquired Conditions measures. The actual rates of falls with trauma occurring within the hospital may be found on the Medicare data Web site (use filter to restrict reported measures to "Falls and injuries"): https://data.medicare.gov/Hospital-Compare/Hospital-Acquired-Condition-Reduction-Program/yq43-i98g.

5.1.8. How can you improve the quality of the data being collected for fall rates?

To improve data quality, you will need to improve staff reporting of falls, particularly the circumstances surrounding the fall (go to Tool 3O, "Postfall Assessment for Root Cause Analysis"). Often, critical details are left out in the reporting of falls and there are only limited opportunities to learn what makes for a good incident report. Therefore, consider reviewing completed incident reports with staff on a monthly basis.

Check on the quality of the incident reports being filled out at your hospital or on your unit using Tool 5A, "Information To Include in Incident Reports".

Return to Contents

5.2. How do you measure fall prevention practices?

5.2.1. Why measure fall prevention practices?

While measuring fall rates is the ultimate test of how your facility or unit is performing, fall rates are limited in that they do not tell you how to improve care. If your fall rate is high, on what specific areas should you focus? To know where to focus improvement efforts, it is important to measure whether key practices to reduce falls are actually happening.

Many important practices could be measured in assessing fall prevention. We recommend initially looking at no more than two, such as:

• Performance of fall risk factor assessment within 24 hours of admission.
• Performance of care planning that addresses each risk factor identified during fall risk factor assessment.

5.2.2. How do you review performance of a fall risk factor assessment within 24 hours of admission?

As the first step in prevention, it is essential to ensure that a fall risk factor assessment is performed within 24 hours of admission. The risk factor assessment could either be a standard scale such as the Morse Fall Scale (Tool 3H) or STRATIFY (Tool 3G), or it could be a checklist of risk factors for falls in the hospital.

The key question is not so much whether a scale was used, but rather whether the known risk factors for falls were assessed. In some cases, the risk factors will vary depending on the hospital unit, so the risk factor assessment may need to be tailored to the unit. Determine whether this fall risk factor assessment is being performed.

5.2.3. How do you assess care planning to ensure that it addresses each deficit on the fall risk factor assessment?

For risk factor assessment to make a difference, all risk factors identified on the risk factor assessment need to be addressed in the care plans, and the care plans need to be acted on. This requires critical thinking on the part of staff and a tailored approach to each patient based on the individual patient's risk factors. Ensure that the care plans address all areas of risk.

Find detailed instructions on how to perform a review of medical records at the Duke University Medical Center Patient Safety/Quality Improvement Web site: http://patientsafetyed.duhs.duke.edu/module_b/ module_overview.html .

5.2.4. What data sources should be used in measuring fall prevention practices?

In measuring key practices, data used in calculating performance rates can be obtained from a number of sources. These include direct observations of care, surveys of staff, and medical record reviews. Each approach has its strengths and limitations:

• Direct observation of care, where a trained observer determines, for example, whether a patient's call light is within reach, will be the most accurate approach for certain care processes but can be time consuming.
• Surveys may be helpful in certain circumstances but rely on staff members' recall of specific events, and these recollections might be inaccurate.
• Medical record reviews are the easiest approach to complete but rely on what is documented in the record, and much care for fall prevention may not be documented.

As a starting point, we recommend that you combine medical record review with direct observation using a manageable sample size (e.g., no more than 20 patients), as suggested in Tool 5B.

Use this tool adapted from the Royal College of Physicians FallSafe program for auditing key processes of care (Tool 5B, "Assessing Fall Prevention Care Processes").

5.2.5. What should be done if you are not doing well on your measures of fall prevention practices?

Good performance on these key processes of care is critical to preventing falls. If you are not doing well, or as well as you would like, in one of these key areas, it provides an opportunity for improvement. Examine what the problem is and plan how to overcome this barrier. For example, are staff engaged in the program? Do they know what they need to do? Go back to section 2.2 for suggestions on how to make needed changes.

Return to Contents

5.3. Checklist for measuring progress

The checklist for measuring progress can be found in Tools and Resources (Tool 5C, "Measuring Progress Checklist").