The Pennsylvania Patient Safety Authority set out to evaluate the interest of Pennsylvania hospitals to expand a regional initiative to standardize falls reporting to a statewide level. A survey was distributed to Pennsylvania hospitals to determine the level of interest in this initiative. The results showed a general consensus among 81respondent hospitals on the standardized falls definitions and willingness to participate. Hospitals indicated support for separate reporting of comparison data of falls rates for general acute care and specialty hospitals and for inpatient and outpatient settings. Potential barriers include financial, resource, and data collection burdens (e.g., collecting unit-level utilization data).
Beginning in the fourth quarter of 2008, the Pennsylvania Patient Safety Authority partnered with the Health Care Improvement Foundation (HCIF) and 29 hospitals in the southeast region of Pennsylvania to standardize reporting of patient falls. At the end of the fourth quarter of 2010, this initiative culminated in a patient falls conference for participating hospitals to share their experiences. Given the positive response to the regional initiative, the Authority and the Hospital and Healthsystem Association of Pennsylvania explored the opportunity to expand this initiative statewide. Input from various hospitals led to the development of a survey to ascertain the potential to expand this project. The purpose of this article is to summarize the results of this statewide survey.
In August 2011, the Authority surveyed Pennsylvania hospitals to evaluate statewide expansion of the initiative to standardize reporting of patient falls. The survey was sent to 167 general acute care hospitals, specialty hospitals, children’s hospitals, and rehabilitation hospitals or centers. Invitations to 13 facilities were undeliverable, which reduced the sample size to 154 facilities. The survey included 25 questions: 18 general questions, 5 follow-up questions, and 2 open-ended questions. When analyzing the results, the analysts interpreted nonresponses in two ways. For questions of preference, the analysts interpreted nonresponses as indications of no preference and included them in the analysis. For all of the other questions, the analysts removed nonresponses from the analysis and calculated percentages based on specific responses. The unit of analysis is the hospital.
The survey response rate was 52.5% (n = 81 of 154). Among respondents, two questions had a 100% response rate (n = 81 of 81), the remaining response rates ranged from 69% to 94%. General acute care hospitals made up the majority of respondents (80%). The Figure shows a breakdown of the respondents. Figure. Falls Survey Hospital Distribution
Hospital Comparison Data
Questions about falls comparison data focused on current involvement of hospital comparison of falls rates with external data sources, participation in a Pennsylvania statewide standardized reporting of falls rates initiative using comparison data, and which outpatient services and departments to include in the reporting of comparison data.
External data sources. Seventy-nine percent (n = 64 of 81) of responding hospitals compared their falls rates with an external data source; 21% (n = 17 of 81) did not. Eleven different types of external source comparison data were identified and used by this sample of Pennsylvania hospitals. Some hospitals identified multiple external data sources; however, the National Database of Nursing Quality Indicators (NDNQI) was the biggest external source (59.4%). Table 1 provides a breakdown of the external data sources.
Statewide comparison data. The majority of hospitals, 85.2% (n = 69 of 81), were interested in participating in a Pennsylvania statewide standardized reporting initiative that would compare falls rates and improve processes, while 8.6% (n = 7 of 81) were not interested and the remaining 6.2% (n = 5 of 81) indicated no preference. A similar level of support was indicated for the ways hospitals prefer to have comparison data presented. There was strong support for separate reporting of comparison data of falls rates for general acute care and specialty hospitals and for separate reporting of comparison data of falls rates for inpatient and outpatient settings. See Table 2.
Types Of Separate Standardized Reporting
Of Comparison Data
|Prefer Separate Standardized Reporting |
Of Comparison Data
Of Falls Rates
|Do Not Prefer Separate Standardized Reporting Of Comparison Data Of Falls Rates||No Preference For Standardized Reporting Of Comparison Data Of Falls Rates*|
|Acute care general hospitals versus specialty hospitals||68 (84%)||0 (0%)||13 (16%)|
|Inpatient versus outpatient settings||62 (76.5%)||5 (6.2%)||14 (17.3%)|
|* No preference reflects nonresponses to these specific survey questions |
Outpatient comparison data. Hospitals were asked which outpatient settings/departments for falls rates comparison data would be desired. They identified nine different areas, with five settings/departments receiving 97% of the total responses. See Table 3.
|Outpatient Settings/Departments||Total Responses*||Percentage|
|Ambulatory surgical procedural units||3||4.5|
|Postanesthesia care unit||1||1.5|
|Nonclinical buildings and grounds||1||1.5|
|* Out of 67 responding hospitals. Facilities provided more than one response.|||||
The Authority/HCIF southeastern regional falls reporting initiative defined a fall as “any unplanned descent to the floor (or other horizontal surface such as a chair or table), with or without injury to the patient.”
This definition of falls includes assisted falls, in which a caregiver sees a patient about to fall and intervenes, lowering him or her to a bed or floor, and therapeutic falls, in which a patient falls during a physical therapy session with a caregiver present specifically to catch the patient in case of a fall. The definition excludes failures to rise, in which a patient attempts but fails to rise from a sitting or reclining position.
The majority of responding hospitals, 80.5% (n = 58 of 72), used the falls definition identified by the initiative, and 19.4% (n = 14 of 72) of responding hospitals did not. When asked if changes in the definition should be made, 70.8% (n = 51 of 72) of responding hospitals said no. The remaining 29.2% (n = 21 of 72) of responding hospitals provided comments or recommended changes to the definition. Two respondents recommending changes indicated that they would change their facilities’ falls definition to the falls definition identified by the initiative. The recommended changes were grouped according to 10 similar themes and are presented in Table 4.
|Recommended Changes Or Comments||Number Of |
|Percentage Of Respondents|
|Therapeutic and developmental falls are not falls.||6||28.5%|
|Assisted falls are not true falls.||4||19.0|
|Definition is not appropriate for behavioral health and rehabilitation facilities.||3||14.3|
|No recommended changes.||3||14.3|
|Hospitals would change their definition to the falls definition in the survey.||2||9.5|
|Failure to rise is ambiguous and inconsistent to measure.||2||9.5|
|The use of Steri-Strips or glue should be considered first aid.||1||4.7|
|Add falls risk assessment, patient population, and presence of staff to definition.||1||4.7|
|Provide distinction for anticipated versus unanticipated falls.||1||4.7|
|There are discrepancies between the initiative's definition and the National Database of Nursing Quality Indicators' definition, which includes the use of high or low beds.||1||4.7|
|* From 21 responding hospitals. Some respondents provided more than one comment. |
Hospitals were asked whether they use the defined falls event subcategories (e.g., toileting, ambulating) in the Pennsylvania Patient Safety Reporting System (PA-PSRS) or whether they use the write-in description subcategory labeled “Other” when reporting falls data. A majority of hospitals, 74.6% (n = 53 of 71), responded that they use the PA-PSRS falls event sub-categories; however, 25.4% (n = 18 of 71) of hospitals responded that they report falls using the write-in falls subcategory “Other.” One-half (n = 9 of 18) of the hospitals that use the write-in subcategory “Other” to report falls events did not provide a reason for using this subcategory. Twenty-seven percent (n = 5 of 18) of these hospitals identified patient conditions (e.g., seizures, syncope) as a falls event subcategory, and the remaining 22% (n = 4 of 18) of these hospitals identified other circumstances of the fall (e.g., found on floor, intentional falls).
When hospitals were asked about the falls with harm definition requiring the inclusion of any fall where more than first aid care is needed, 94% (n = 66 of 70) of respondents agreed with the definition. Three percent (n = 2 of 70) of hospitals recommended the removal of first aid and minor sutures from the current falls with harm definition, and another 3% (n = 2 of 70) of hospitals inquired about how to classify rib fractures that do not require surgical intervention.
Collection and Measurement Issues
General data collection concerns. Data collection questions assessed potential reasons for nonparticipation and current data submission practices. More than half, 59.2% (n = 45 of 76), of the responding hospitals would participate in the standardized patient falls reporting initiative even if additional data collection were required. The remaining hospitals, 40.8% (n = 31 of 76), were uncertain or would find additional data collection prohibitive. The most common reasons for lack of participation included resource limitations (financial, time, and personnel) and data collection burden, especially being required to collect new types of data in addition to current data.
Measurement issues. Responding hospitals almost unanimously (94.3% [n = 66 of 70]) acknowledged that they collect and monitor falls at the unit level, yet only 77.9% (n = 53 of 68) of hospitals were willing to report patient-days separately for every unit. Potential methods or resources to consider should the initiative require unit-level data include the following:
- Use of NDNQI data methods and definitions
- Use of PA-PSRS, specifically the patient-days reporting component of the infection control system
- Use of 1,000 patient-days as the standard denominator for reporting falls
- Use of a web source to report data or excel spreadsheet to collect data for reporting
There are several limitations to this survey. It was sent to a sample of Pennsylvania hospitals with a response bias toward acute care general hospitals and possibly hospitals actively involved in reporting performance measures and adverse event data. Distributing the survey in late August for a period of two and a half weeks may have influenced the number of responses. Survey responses may be influenced by respondents’ involvement in the recent Authority falls reporting initiative or other negative or positive reporting experiences.
This sample of Pennsylvania hospitals indicates a willingness to participate in a statewide initiative to standardize reporting of falls rates as well as a general agreement on the falls definition. Overall, the consensus indicates that Pennsylvania hospitals actively compare falls rates to a variety of external data sources and want more facility-level, unit-level, and inpatient- and outpatient-specific information. There are several common themes to the issues of data collection and measurement, namely the standardization of definitions and data collection requirements and a desire to use established systems (e.g., PA-PSRS, NDNQI, Centers for Medicare and Medicaid Services or Institute for Healthcare Improvement reporting systems) to facilitate reporting of falls rates. Incorporating the survey information into a statewide standardized falls reporting initiative would expand and enhance hospital monitoring and may improve understanding of patient falls.
Theresa Arnold, DPM, Pennsylvania Patient Safety Authority; Denise Barger, BA, CPHRM, CPHQ, CPPS, HEM, Pennsylvania Patient Safety Authority; and Pamela A. Braun, RN, MSN, Health Care Improvement Foundation, developed the hospital statewide falls survey.