Monitoring falls in residential aged care facilities: Agreement between falls incident reports and progress notes

Accurate fall reporting is essential for assessing the effectiveness of fall prevention strategies. This study aimed to investigate the level of agreement between incident reports and resident progress notes as data sources for falls monitoring in residential aged care facilities.


| INTRODUCTION
The global health issue of falls in older population affects 27-34% of community-dwelling older people annually. 1In residential aged care facilities (RACFs), the risk of falling is twice as high as in the community, leading to increased morbidity, mortality and loss of functional independence. 2,3The annual cost of healthcare services related to fall is approximately $20 billion in the United States 3 and $4.3 billion in Australia. 4ccurate fall data are crucial for assessing the impact of fall prevention strategies.Falls can be collected retrospectively in which the individual is expected to recall their fall history over a time period or prospectively via progress notes or incident forms. 5In Australia, mandatory fall reporting was established in 2021 through the National Aged Care Mandatory Quality Indicator program. 6Online incident reports are commonly used to describe fall rates at a facility level, providing quick results compared to manual audits of progress notes. 7However, discrepancies between hospital falls data in incident reports and progress notes highlight the need to investigate their agreement in RACFs. 8his study aimed to assess the agreement between fall incident reports and progress notes in RACFs and explore potential variations in agreement across different aged care facilities.This study sought to determine whether incident reports alone can reliably and effectively identify falls in RACFs, summarising information on the incidence, characteristics and outcomes of falls over a 12-month period.

| Study design
This was a quantitative study investigating the level of agreement between progress notes and incident reports to monitor falls data in RACFs.This study was conducted within the TOP UP trial, a randomised controlled trial that aims to examine the effectiveness of telehealth physiotherapy and exercise on mobility performance and falls for older people in aged care.(ANZCTN:12621000734864).

| Setting
The first six RACFs that were recruited to the TOP UP trial were included in this study.One metropolitan site in Sydney in the Australian state of New South Wales-Facility A, and five rural sites across New South Wales (Facilities B to F).All sites delivered high-level dependency and dementia care.

| Participants
All TOP UP study participants from the six RACFs were included in this study.Eligibility included older people (65 years and over), possessing sufficient neurological, cognitive and sensory skills to participate in the program with individual written consent or consent from their person responsible.Those with terminal or unstable illness, severe dementia, who had participated in a similar physiotherapy program in the last year or who were unable to walk 10 m were excluded from the study.

| Data collection
An Excel data extraction sheet was developed by the research team and pilot-tested on the first 10 participants.Authors (Author 1 and Author 2) made modifications based on pilot results to ensure all relevant data were collected including length of stay.One author (Author 2) remotely accessed online incident reports and progress notes to extract fall data (fall date, time, activity, equipment involved, injuries, hospital admission and follow-up) and participant characteristics (age, gender, cognitive status, comorbidities and mobility).The information was collected from available data 12 months before randomisation to the TOP UP trial, which commenced in September 2021.If data were unavailable for the preceding 12 months, the surveillance period was recorded in months.The time it took to collect data from the two incident methods was also collected.The audits used the World Health Organization's (2021) definition of a fall, which is an event resulting in a person

Practice Impact
This study demonstrated substantial agreement between falls in residential aged care recorded via incident reports and progress notes.It appears that incident reports can be used as a convenient but valid method for monitoring falls data, in which staffing is sufficient to allow for the effective use of incident reporting systems.
inadvertently coming to rest on the ground or a lower level. 9Injurious falls (skin tears, head injuries or fractures) were also recorded.
This project had ethics and governance approval by the Human Research Ethics Committee at The Sydney Local Health District, Concord, Australia (approval number approval CH62/6/2021-009).

| Data analysis
Participant and fall characteristics were presented using descriptive statistics (number and percentage where applicable).The sum of independent falls captured by either the incident reports or the progress notes was calculated.The level of agreement between the two data collection methods for the total sample and for each of the six facilities was assessed by Cohen's kappa (κ) coefficient, in which values of 0 would indicate no agreement, 0.01-0.20 as slight, 0.21-0.40 as fair, 0.41-0.60 as moderate, 0.61-0.80 as substantial and 0.81-1.00as almost perfect agreement. 10IBM SPSS Statistics version 28 (IBM Inc., USA) was used to analyse the data.

| Participant characteristics
Among 420 residents living in six RACFs, 46 participants were included.The participants had a mean age of 83 years (standard deviation [SD] 9), with 65% being female (n = 30).Of these, 78% (n = 36) had a cognitive impairment, including 28 with mild impairment.On average, participants had six comorbidities and 74% (n = 34) required a walking frame for mobility.Reviewing progress notes took 1-2 h, while incident reports took approximately 5 min.Additional participant characteristics are in Table 1.
There were 75 falls recorded using either incident reports and/or progress notes with a mean falls' incidence of 1.84 falls per person-year.Twenty-seven participants (59%) fell.Most falls (39 out of 75, 52%, 3.89 falls per personyear) were recorded from Facility B. Falls were less common during the night (12 a.m.-5:59 a.m.) with the other times during the day showing similar fall rates.Falls most occurred in the resident's room (45%) or their bathroom (17%) and occurred when walking (37%) or transferring (36%).Forty-two (56%) of fall incidents were recorded by a registered nurse and 31 (41%) were recorded by a care staff assistant.Further fall characteristics are presented in Table S1.

| Level of agreement
The agreement level (kappa value) is displayed in Table 2. Sixty-six out of 75 (88%) falls were captured in both methods, with seven falls found only in the progress notes and two falls found only in the incident reports.Incident reports identified 68 falls (90.7%) whilst progress notes recorded 73 falls (97.3%).The level of agreement between incident reports and progress notes across all facilities was classified as a substantial agreement (0.75; 95% CI 0.59 to 0.90).We found perfect agreement for all facilities (κ = 1.00), except for Facility B, classified as a fair agreement (0.29; 95% CI −0.05 to 0.62).

| DISCUSSION
Accurate reporting of falls is crucial in RACFs to identify risk, implement preventive measures and enhance the safety of older people.The results of this study demonstrate a substantial agreement between monitoring falls data in RACFs from incident reports and progress notes.Both methods identified most falls.All facilities achieved perfect agreement except Facility B. Workforce shortages could explain this lack of agreement at Facility B. Given the labour-intensive nature of collecting fall data from progress notes, these findings suggest that incident reports can be used as an accurate and more efficient method of fall data monitoring for research purposes.
The incidence rates and fall characteristics of this study are consistent with previous research in RACFs. 2,11A recent epidemiological study of falls in Australian RACFs found that 58% of residents experienced falls 9 and, in Germany, the mean fall incidence rate was 1.74 falls per person-year. 2Both studies showed that the highest risk of falls was associated with residents unsupervised in their bedrooms.Future research that can lead to better monitoring of residents in their bedrooms is warranted to help inform future fall prevention strategies.
Facility B exhibited the highest rate of falls (n = 39) and exhibited only fair agreement between the two fall collection methods.The facility also had the highest incidence of repeat non-injurious people who fell, with five residents experiencing three or more falls.During the study, this facility sustained workforce challenges, including difficulties employing registered nurses due to rural location and sustained sick leave exacerbated by repeat COVID-19 outbreaks, which may have contributed to the discrepancies in accurate fall data collection.
Previous research suggests nursing staff tend to underreport non-injurious falls especially when they perceive to be time poor. 12To address this issue, further research involving mixed methods is recommended to understand staff workload and attitudes towards fall reporting, enabling the development of targeted interventions for resident safety. 13hile both methods of fall reporting are used in aged care, they offer differing levels of efficiency and convenience. 7Progress notes are an important tool as they allow registered nurses to communicate the fall incident to all care staff and allow for immediate fall follow-up and care plan evaluation. 14However, incident reports have shown to be much more time-efficient and offer a comprehensive review of falls than progress notes. 15

| CONCLUSIONS
This study highlights that there is substantial agreement between incident reports and progress notes with perfect agreement found in all facilities except for one site.Incident reports offer a convenient method for fall monitoring for clinical and research purposes in RACFs compared to progress notes.However, the discrepancies between the two fall reporting methods exist, highlighting Participant characteristics.Agreement between falls reported in progress notes and incident reports.