Cost of Inpatient Falls and Cost-Benefit Analysis of Implementation of an Evidence-Based Fall Prevention Program

Key Points Question What are the costs of inpatient falls and cost benefits associated with the Fall TIPS (Tailoring Interventions for Patient Safety) Program? Findings In this economic evaluation using a large cohort (900 635 patients; 7858 noninjurious falls; 2317 injurious falls), the average total cost of a fall was $62 521 ($35 365 direct costs), and injury was not significantly associated with increased costs. The Fall TIPS Program was associated with $22 million in savings at study sites across the 5-year study period. Meaning The findings of this study indicate that implementation of cost-effective, evidence-based safety programs was associated with lower cost and care burdens associated with inpatient falls and are a step toward safer, more affordable patient care.

Many hospital falls can be prevented through implementation of an evidence-based program that identifies each patient's fall risk factors, develops individualized prevention plans, and consistently implements the plans through staff and patient engagement. 8,13,14However, adoption of such programs is limited.The purpose of this study was to use electronic health record (EHR) data to estimate the cost of falls and related injuries and to analyze the costs and benefits associated with implementing the Fall TIPS (Tailoring Interventions for Patient Safety) Program, an evidence-based, freely available fall prevention program associated with a 15% to 25% reduction in inpatient falls and a 0% to 34% reduction in injurious falls. 8,15 assessed the costs of inpatient falls (2021 US dollars) before, during, and after implementation of the Fall TIPS Program across 2 large health care systems.We categorized fall severity on a scale from noninjurious to severe or death 16 to understand how degrees of injury are associated with costs.Finally, we assessed the cost benefits associated with program implementation.

Study Design and Participants
In this economic evaluation, we performed a cost-benefit analysis of implementing the Fall TIPS Program with the primary outcome of cost of inpatient falls.A secondary analysis quantified the costs and savings associated with the evidence-based fall prevention program.Our base model estimated the total cost savings of intervention effects (ie, reduction in overall direct and total costs of hospital stay).The model was framed from the perspective of the health care system, and data on costs and outcomes were obtained from a nonrandomized interrupted time series (ITS) study conducted across 2 large health care systems in the Bronx, New York (site 1; 3 hospitals), and Boston, Massachusetts (site 2; 5 hospitals).All hospitals implemented the Fall TIPS Program on medical and surgical units.The ITS evaluation was conducted between June 1, 2013, and August 31, 2019, to evaluate the Fall TIPS Program's outcomes and compared the falls and fall injury rates (eAppendix and eFigures 1 and 2 in Supplement 1).We also conducted a case-control study to estimate the additional direct and total costs associated with fall and injury level.This study followed the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) and Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines.The institutional review boards for Montefiore and Mass General Brigham Healthcare Systems approved the study.
Due to the quality-improvement nature of the intervention, a waiver of informed consent was granted by the institutional review boards of Montefiore and Mass General Brigham Healthcare Systems.

Study Design and Intervention
All sites had a 15-month preintervention period from June 2013 to September 2014.In collaboration with hospital leadership, the study team assigned the go-live period between September 2014 and May 2018 based on EHR go-live dates and other competing projects (Figure).All hospitals implemented the Fall TIPS risk assessment and care planning tools in their EHR.Boston sites also implemented the Fall TIPS laminated poster. 8,14New York sites implemented the EHR-generated Fall TIPS poster. 8,14Both modalities are effective in facilitating patient engagement in fall prevention. 17otocol adherence was measured with patient engagement audits conducted by unit-based champions. 8All sites completed data collection in August 2019.

Outcomes
The primary outcomes were the total and direct costs of falling compared with not falling.The secondary outcomes were the costs and cost savings associated with the evidence-based fall prevention program.

Case-Control Study
We assessed the costs of falling compared with not falling.Cases were defined as all patients who were reported as falling in the incident reporting system of participating hospitals from June injury, and severe injury with log of cost, a continuous dependent variable (using the log transformation because cost was right skewed), using random-effects models to account for matching.We then reexamined these associations using random-effects regression to control for potential confounding due to clinical and socioeconomic factors.Data analysis was performed from October 2021 to November 2022.We used SAS statistical software, version 9.4 (SAS Institute) for the analyses.

Cost-Benefit Analysis
We leveraged the findings of the case-control study to determine costs and benefits associated with implementing the Fall TIPS Program.Includes a 2-mo integration period prior to the postintervention period a Site 2A (consisting of 26 units within 1 hospital) staggered the postintervention period.All other hospital sites used the same start and end dates for all units.Roman numerals indicate the grouping of when each unit started the postintervention period.

Costs of Implementation
paper posters for each patient at an estimated cost of $1.00 per patient.Ongoing costs may be incurred because of additional time spent using the Fall TIPS Program.A survey indicated a majority of nurses (62%) averaged 1-minute additional time each day per patient, while 38% reported averaging a 10-minute time savings each day per patient. 18Because of the broad range of results, we assume that the Fall TIPS Program is time neutral.

Benefits of Implementation |
The benefits attributable to the Fall TIPS Program included any reduction in patient falls and injurious falls after implementation.We value the benefits as the avoided costs of evaluating and treating a patient who had a fall event.We obtained figures for the average cost of a hospital stay for a patient with (1) no fall, (2) a fall but no injury, and (3) a fall with any injury.This narrowly focused estimate of benefits does not include the pain and suffering avoided by the patient, nor the opportunity cost of the patient's time had they required a longer LOS.

Cost-Benefit Analysis |
We compared the costs and benefits of the Fall TIPS Program to evaluate a snapshot of the costs and benefits at the start and at the end of the intervention periods.

Study Design and Participants
A total of 900 635 patients and 4 955 534 patient-days were included in the ITS (Table 1; eAppendix 3) days to 5.6 (7.9) days, and the mean (SD) unit LOS ranged from 4.4 (6.1) days to 4.6 (6.0) days.There were more women in each period than men (52.3%-53.6%women vs 46.5%-47.7%men).More patients were younger than 65 years in all periods (51.1%-56.7%patients were <65 years; 43.3%-48.9%patients were Ն65 years).Standardized differences comparing demographic characteristics across periods were well balanced (<10%), except race and ethnicity when comparing the preintervention period with the go-live period and the total Charlson Comorbidity Index score at admission when comparing the preintervention period with the postintervention period.We adjusted for these variables in the analyses (see eTables 1 and 2 in Supplement 1 for site-specific demographic characteristics).

Case-Control Study
During the 74-month study period, there were 7858 noninjurious and 2317 injurious falls.Table 2 includes the descriptive statistics.The average total cost of a fall was $62 521 ($36 776 direct costs), and the average total cost of a fall with any injury was $64 526 (Table 3).The intervention cost $267 700 (both health care systems), equivalent to $0.88 per patient or $180 per 1000 patientdays.The intervention prevented 567 falls (142 with injury and 425 without injury), resulting in avoided total costs per 1000 patient-days of $14 762 (approximately $8500 in direct costs per 1000 patient-days) (Table 4) in the postintervention period.The net avoided costs per 1000 patient-days totaled $14 600 (approximately $8300 direct costs) for total cost savings of $22 036 714.
Assuming estimates of 25.5 thousand medical or surgical discharges with 123 130 000 annual patient-days nationally 19 and extrapolating cost savings from this intervention, we project annual cost savings of $1.82 billion (direct cost savings $1.05 billion) set against projected total intervention costs of $20 million.We used the national average registered nurse hourly wage ($39.78) 20for this

Discussion
We found that the average total cost of a fall was $64 526 ($36 776 direct costs) and that the level of injury was not significantly associated with cost.The Fall TIPS Program was associated with a total   [10][11][12] and other hospital-based fall prevention program evaluations demonstrate mixed cost-effectiveness results, in which the costs of some programs were greater than potential savings. 22A 2016 report 23 contracted by the Agency for Healthcare Research and Quality estimated that the cost of a fall (any injury) was $6694 (2015 US dollars) based largely on a case-control study of an inpatient sample (62 cases [ie, patients who had a fall event]) and manual medical record review. 10This same report estimated that for every 1000 falls, there are 50 excess deaths.This present study used actual cost data from the EHR systems of 2 large health care systems (10 176 cases [ie, patients who had a fall event]) and determined the direct cost of a fall with any injury to be $36 776 (total cost $64 526).We stratified falls by severity, included matched controls, and aimed to provide health care leaders with information demonstrating the costs of falls and the benefits of implementing an evidence-based program.The ITS (eAppendix and eFigures 1 and 2 in Supplement 1) included a total sample of 900 635 patients and more than 74 months of data to assess how implementation of the program was associated with costs.The Fall TIPS Program saved $22.0 million in the postintervention period across 2 health care systems ($6.4 million and $15.6 million, respectively) and prevented 50 excess deaths. 23e costs of falls with or without injury were not appreciably different.This finding suggests that even in the absence of obvious injury, postfall evaluation and testing are extensive, and LOS is prolonged.Therefore, programs that prevent all falls provide the greatest cost-savings opportunities.
To our knowledge, this is the largest study to date evaluating the cost of hospital falls, and it builds on existing literature demonstrating the cost-effectiveness of evidence-based fall prevention programs. 8,15,18 performed sensitivity analyses to assess variation in costs on the net benefits of the intervention.Because the material costs averaged approximately $0.88 per 1000 patient-days, we focused on uncertainty in registered nurse time.Based on a previous study, we judged the intervention to be time neutral for nurses. 18If instead we assume the intervention costs nurses an additional 2 minutes per shift per patient-day, then with 1.492 million patient-days in the analysis, the Fall TIPS Program increased costs by $6.88 million.This reduces the total direct cost savings per 1000 patient-days from $14 500 to $10 000 (from $8500 to $3715 for direct costs).In addition, we performed a "break-even" analysis of nurses' time and calculated that an additional 179 registered nurse hours (including nurse champions) per 1000 patient-days could be spent on the Fall TIPS Program before the costs exceeded benefits.This is equivalent to an additional 10.75 minutes per patient each day.Thus, we conclude that the Fall TIPS Program results in net cost savings over a wide range of assumptions concerning nursing time.
In 2008, the Centers for Medicare & Medicaid Services (CMS) ended fall-related cost reimbursement, 24 a controversial policy because some falls are not preventable. 25Many hospitals responded by implementing fall prevention strategies supported by little or no evidence. 26Today there is wide variation in the implementation of effective fall preventive strategies.Financial incentives within the national quality payment program have been used to decrease the frequency and cost of patient falls, but, to date, they address only fall injuries and a minority of cases.For example, the 2008 Inpatient Prospective Payment System initiative enacted to prevent hospitalacquired conditions (HACs) mandated that the CMS no longer pay for conditions that (1) were high cost or high volume, (2) resulted in higher payment when present as a secondary diagnosis, and (3) were considered preventable. 27Under the final rule, HACs are identified through claims data.
Hospitals are required to report present on admission (POA) information status for principal and secondary diagnoses when submitting claims.This study analyzed the costs and benefits of preventing falls using the Fall TIPS Program from the health care system perspective.Findings can be used to assist other organizations in evaluating the decision to invest in implementing an evidence-based fall prevention program.Findings can also be instructive from a public policy stance as this program is beneficial for patient safety, results in cost savings, and uses validated materials that are available free of charge in 9 languages. 32Resources to improve patient safety are limited, and the benefits associated with the Fall TIPS Program far outweigh the associated costs.

Limitations
We estimated total savings associated with the reduction in costs of fall-related care from the perspective of health care organizations.Data on costs were obtained through an ITS study in 2 large health care systems, and indirect costs were available in less precise forms.We did not have access to a breakdown of the components of direct and total costs; we expect that a major contributor to increased costs following a fall event is increased LOS.Our ITS design did not include a control series, so we cannot exclude confounding from co-occurring interventions or changing hospital dynamics that may have impacted hospital fall rates.
A case-control study with the same cohort was used to estimate costs based on fall injury level.

Table 1 .
Patient Demographic Characteristics (Sites 1 and 2) Across periods, patients were similar in hospital and unit LOS, age, sex, and insurance type.Patients at site 1 were more likely to be non-White than White (79.8%-82.6%non-White[American Indian or Alaska Native, Asian or Pacific Islander, Black or African American, Native Hawaiian or Other Pacific Islander] across periods), whereas patients at site 2 were more likely to be White than non-White (82.1%-83.3%Whiteacross periods).The mean (SD) hospital LOS ranged from 5.4(7. Abbreviations: EHR, electronic health record; NA, not applicable.aConsideringstandard differences of less than 10% as not significant.bRace data are based on self-reported EHR data.Racial and ethnic categories in the non-White group include American Indian or Alaska Native, Asian or Pacific Islander, Black or African American, Native Hawaiian or Other Pacific Islander.Categories in the White group include White or Caucasian.cTen percent of ethnicity data missing across sites.JAMA Health Forum | Original Investigation Inpatient Falls and Implementation of an Evidence-Based Fall Prevention Program JAMA Health Forum.2023;4(1):e225125.doi:10.1001/jamahealthforum.2022.5125(Reprinted)January 20, 2023 4/11 Downloaded From: https://jamanetwork.com/ on 09/23/2023and eFigures 1 and 2 in Supplement 1).

Table 3 .
Cost Table

Table 4 .
Cost Savings cost savings of $22 million over approximately 5 years at the intervention sites, projected to a nationwide annual cost savings of $1.82 billion.Information on the cost of inpatient falls is limited, outdated, and variable, Abbreviations: ITS, interrupted time series; NA, not applicable.a Number of patients in the postintervention period.b Estimated number of patients, model-based estimate based on the ITS.JAMA Health Forum | Original Investigation Inpatient Falls and Implementation of an Evidence-Based Fall Prevention Program JAMA Health Forum.2023;4(1):e225125.doi:10.1001/jamahealthforum.2022.5125(Reprinted) January 20, 2023 6/11 Downloaded From: https://jamanetwork.com/ on 09/23/2023

on 09/23/2023 resulted
31sed on fiscal year 2011 data, 1 report found approximately 89.3 million secondary diagnosis claims, but more than 75% were reported as POA.The Falls and Trauma HAC category were the most frequently reported secondary diagnoses, but only 3.2% were coded as not POA.While only a small minority of fall injuries were coded as POA, the Falls and Trauma category contained the greatest percentage of hospital discharges (27.2%) that in a reassignment of Medicare Severity-Diagnosis-Related Groups28and hospitals absorbing the cost.The CMS implemented the HAC Reduction Program 29 to link Medicare payments to health care quality.Under that program, hospitals that rank in the worst-performing quartile receive a 1% reduction in overall Medicare payments.The total HAC score includes the CMS Patient Safety and Adverse Events Composite (CMS PSI 90), which consists of a weighted average of In-Hospital Fall with Hip Fracture Rate and 9 other HACs.Most major fall injuries, which range in severity from those that cause temporary functional impairment (ie, dislocated shoulder or broken teeth) to injuries associated with increased mortality (ie, skull fractures and subdural hematomas), are not included in this measure.Recently, the CMS announced that it plans to suppress the HAC Reduction Program payment penalties for fiscal year 2023 due to the impact of the COVID-19 public health emergency on data reporting efforts during the pandemic.30Participatinghospitalswillnot receive a total HAC score or payment penalty, but the CMS will calculate and publicly report the CMS PSI 90 scores.This pause in CMS penalty-based reductions may provide an opportunity for hospitals and health systems to reevaluate their fall prevention programs and to adopt an evidence-based program.While the CMS has implemented policy disincentives to reduce fall injuries and has created a "never event" designation for inpatient falls,31it has not promoted an evidence-based fall prevention tool.Data from this study suggest that policies that incentivize hospitals to prevent all falls may be the most cost-effective.The CMS should promote evidence-based fall prevention programs like the Fall TIPS Program.This cost-benefit analysis is based on academic medical centers and community hospitals and should be generalizable to other organizations using the Fall TIPS Program.The level of detail to Downloaded From: https://jamanetwork.com/ which costs can be compared depends on hospital-specific cost differences.This analysis could have accounted for true one-time development costs, but other hospitals will not incur development costs because the Fall TIPS Program already exists.Conversely, costs incurred and cost benefits depend on existing organizational structures to support patient safety.Cost benefits can be extrapolated by scaling results according to the number of patients.It may be possible to use data on patient characteristics, falls or injuries, and safety culture scores to perform a more sophisticated extrapolation to typical patient populations in other types of hospitals.
To assess training costs, we multiplied the duration of training by the weighted average hourly wage for nurses by state, 21 by the number of nurses in each health care system.While costs of direct training are calculable, the cost of time associated with the nurse champion responsibility of day-to-day training and advising fellow nurses is unknown.We included 60 minutes of general training to account for the initial 30-minute training plus 30 minutes of follow-up and reinforcement.We used $46.06 as the weighted average hourly wage for nurses.The mean hourly wage for JAMA Health Forum | Original Investigation Inpatient Falls and Implementation of an Evidence-Based Fall Prevention Program SUPPLEMENT 1. eAppendix.Interrupted Time Series (ITS) Methods and Results eFigure 1. Results of the Adjusted ITS Analysis for Overall Fall Rate per 1000 Patient Days Across the Three Periods eFigure 2. The Results of the Adjusted ITS Analysis for Overall Fall Injury Rate per 1000 Patient Days Across the