Back to Evidence

Morse-Tiered Individualized Fall Prevention in Older Inpatients: Evidence from a 365-Patient Prospective Cohort across High-Risk Wards

Evidence-Based Medicine40 min read

Falls are among the most serious safety threats in older inpatients, and uniform prevention often fails because it is not matched to individual risk. This article interprets a 2024 prospective cohort in four high-risk wards of a tertiary hospital: in 365 patients allocated by block randomization, Morse-based tiered care reduced falls from 14.7 to 8.2 per 1,000 bed-days (P=0.003) and severe injury from 11.0% to 4.9%. QSevidence supports grading and evidence checks.

Morse-Tiered Individualized Fall Prevention in Older Inpatients: Evidence from a 365-Patient Prospective Cohort across High-Risk Wards

Best for: Nursing and quality teams in geriatrics, neurology, orthopedics, and cardiology; clinical pharmacists; fall-prevention program managers; and evidence-based nursing researchers. Primary keywords: older inpatients; falls; Morse Fall Scale; risk stratification; multidisciplinary ward cohort; night-time toileting; medication review

Abstract / Short Answer

Between January and December 2024, 365 older inpatients from four high-risk wards of a tertiary hospital were allocated 1:1 by block randomization (block size 4, stratified by ward) to an intervention group (n=183) or a control group (n=182). The intervention group received tiered individualized care based on the Morse Fall Scale (low <25, intermediate 25-45, high ≥45), while the control group received routine prevention. Falls fell from 14.7 to 8.2 per 1,000 bed-days (P=0.003, a reduction of about 44%); injury rates after falls were 28.0% versus 42.1% (P=0.02) and severe injury rates 4.9% versus 11.0% (P=0.04). High-risk falls clustered at night (54.5%) and in the bathroom (48.5%), and were closely linked to toileting and rising from bed (72.7%). Multivariable analysis identified prior fall history (OR=3.12), sedative-hypnotic use (OR=3.47), and high-risk Morse grade (OR=3.12) as independent risk factors, with a multiplicative effect between polypharmacy and night-time toileting (interaction P=0.021). The intervention group had shorter stays and better discharge FES-I and Barthel scores.

Background: Fall Burden, Cumulative Risk, and the Limits of Uniform Prevention

According to the World Health Organization, falls are the second leading cause of unintentional injury death worldwide, accounting for about 684,000 deaths each year. In China, fall mortality and disability-adjusted life years related to low bone density have risen continuously since 1990; fall-related mortality in people aged 80 years or older is about 4.71 times that of those aged 60-64. Among inpatients aged 65 years or older, fall rates reach 3%-5%, with injury in 30%-50% of events and fracture as the most serious complication. Falls also trigger a vicious cycle of fear, reduced activity, functional decline, and recurrent falls, prolonging hospital stay and sometimes causing disputes.

Fall risk in older inpatients is multidimensional: internal factors include impaired balance, reduced muscle strength, gait abnormality, cognitive dysfunction, and sensory decline; external factors include slippery floors, poor lighting, polypharmacy (sedative-hypnotics, antipsychotics, antihypertensives, diuretics), and interference from medical equipment. These factors often act cumulatively: an older patient taking sedatives who gets up to toilet at night simultaneously faces drug-induced balance problems, orthostatic hypotension, dim lighting, and the urge to void, so risk rises steeply. Single-dimension prevention is therefore insufficient. Yet the traditional uniform model applies the same warning signs, non-slip footwear, and education to everyone, over-treating low-risk patients while under-protecting high-risk patients; a systematic review found that multifactorial intervention did not significantly reduce falls in geriatric wards (IRR 0.92, 95% CI 0.65-1.30). During protocol design, teams can use the AI guideline retrieval of QSevidence to compare AGS/BGS and NICE recommendations on risk stratification and multifactorial intervention, and use literature evidence appraisal to verify the epidemiologic evidence linking psychotropic drugs to falls, providing a basis for selecting intervention targets.

Methods: A Block-Randomized Prospective Cohort with Three-Tier Intervention

The study ran from January to December 2024 in the geriatrics, neurology, orthopedics, and cardiology wards of a tertiary hospital. It used a prospective cohort design with block randomization (block size 4, stratified by ward) and a 1:1 allocation to routine care or risk-tiered individualized care. The randomization sequence was generated with SAS 9.4 by statisticians not involved in assessment and concealed in opaque envelopes; outcome assessors and data analysts were blinded to allocation. The protocol was approved by the hospital ethics committee and registered in the Chinese Clinical Trial Registry. Based on a routine fall rate of 4.5% and a hypothesized 40% reduction, 1,374 patients per group were required; the study actually enrolled 365 (below the prespecified target, although the primary outcome already reached statistical significance). During the study, teams can use the structured evidence generation of QSevidence to organize eligibility criteria, randomization details, and outcome definitions into a CONSORT/STROBE-aligned reporting checklist that supports methodological transparency.

Step 1: Eligibility criteria and the risk-grading instrument

Inclusion criteria were age 65 years or older, hospitalization of at least 24 hours, ability to complete assessment, and stable vital signs; exclusion criteria were coma or a Glasgow Coma Scale score of 8 or less, acute severe psychiatric illness, expected stay under 48 hours, prior participation in similar research, terminal malignancy, or continuous bed rest. Risk grading used the Morse Fall Scale (six items, 0-125 points: fall history within 3 months, secondary diagnosis, ambulatory aid, intravenous therapy/heparin lock, gait, and mental status); scores of 45 or higher indicated high risk, 25-45 intermediate risk, and below 25 low risk. The scale has been validated in Chinese older inpatients (Cronbach's α=0.78; inter-rater Kappa=0.82). All participating nurses completed standardized training with a post-training Kappa of 0.85 or higher, and a quality controller re-reviewed 10% of cases monthly.

Step 2: Tier-specific intervention measures

The control group received routine care: admission education, bedside warning signs, dry unobstructed floors, night lights, non-slip footwear, accessible call bells, bed rails on request, and rounds every 2 hours. The intervention group added tier-specific measures. Low-risk patients (score <25) received a daily 5-minute individualized education session emphasizing the three-stage bed-exit routine (lie flat 30 seconds, sit 30 seconds, stand 30 seconds), had personal items placed in fixed bedside positions, and were encouraged to walk in the ward twice daily for 10 minutes with a nurse or family member. Intermediate-risk patients (25-45) additionally received a bedside non-slip mat and a bedside commode at night, a fitted four-point cane or walker, medication review by a clinical pharmacist who adjusted the timing or dose of sedative-hypnotics, antipsychotics, antihypertensives, and diuretics, an individualized toileting plan (offering assistance every 2 hours by day and every 3 hours at night), and 24-hour family presence or hourly night rounds. High-risk patients (≥45) had bed rails raised whenever in bed, used a bed-exit sensor pad that triggered a nurses'-station alarm, received careful evaluation before any soft restraint with reassessment every 2 hours, underwent comprehensive assessment by a multidisciplinary team (geriatrician, pharmacist, physical therapist, nurses) within 24 hours, received 30 minutes of daily balance and strength training from a physical therapist, had medication review completed within 48 hours (prioritizing withdrawal or substitution of long-acting benzodiazepines and tricyclic antidepressants), received orientation training three times daily if cognitively impaired or overestimating their ability, and were required to have 24-hour supervision with their names highlighted on the nurses'-station board.

Risk tierMorse scoreTier-specific intervention
Low<255-minute education, three-stage bed-exit routine, fixed item placement, accompanied walking twice daily
Intermediate25-45Non-slip mat and bedside commode, cane/walker, pharmacist adjustment of drug timing and dose, toileting plan, hourly night rounds
High≥45Bed rails, bed-exit alarm pad, careful restraint evaluation, MDT assessment within 24 hours, daily balance training, medication review within 48 hours, orientation training, mandatory 24-hour supervision

Step 3: Outcomes, statistics, and quality control

The primary outcome was the number of falls per 1,000 bed-days, with a fall defined as unintentionally coming to rest on the ground or a lower level, excluding events caused by violence, seizures, or syncope. Secondary outcomes included the fall-related injury rate (events with grade 2 or higher injury), fracture rate, length of stay, and FES-I and Barthel scores assessed within 24 hours of admission and 24 hours before discharge. Analyses used SAS 9.4: continuous variables were compared with t tests or Mann-Whitney U tests and categorical variables with chi-square or Fisher exact tests; fall rates were compared with Poisson regression (offset by bed-days) to estimate incidence rate ratios; risk factors were identified by forward stepwise multivariable logistic regression (odds ratios with 95% CIs). Analyses followed the intention-to-treat principle with sensitivity analysis for missing data. During implementation, nursing teams can use QSevidence to periodically retrieve updated fall-prevention evidence and cross-check intervention measures against recent guidelines and systematic reviews, allowing timely protocol revision.

Key Results: A 44% Reduction in Falls with Fewer Injuries and Severe Injuries

Fall rate and fall-related injury

Baseline characteristics were balanced (mean age 83.9±6.2 years in the intervention group and 83.5±5.9 in the control group; over 70% in both groups had three or more chronic conditions; polypharmacy rates were 68.3% and 66.5%; fall history within 6 months was 32.8% and 33.0%). Falls per 1,000 bed-days were 8.2 (95% CI 6.1-10.5) in the intervention group versus 14.7 (95% CI 11.9-17.8) in the control group (P=0.003), a reduction of about 44%. Injury rates after falls were 28.0% versus 42.1% (P=0.02); severe injury rates (fracture, laceration requiring sutures, or altered consciousness) were 4.9% versus 11.0% (P=0.04). No fall directly caused death in the intervention group, whereas one control patient died of intracranial hemorrhage after a fall.

OutcomeIntervention (n=183)Control (n=182)P value
Falls per 1,000 bed-days8.2 (95% CI 6.1-10.5)14.7 (95% CI 11.9-17.8)0.003
Injury rate after a fall (%)28.042.10.02
Severe injury rate (%)4.911.00.04
Fall-related deaths (n)01-

Temporal, spatial, and activity patterns of falls by risk tier

The intervention group was divided by Morse score into low (<25, n=52), intermediate (25-45, n=78), and high (>45, n=53) tiers. Falls in the high-risk group clustered at night (22:00-06:00), accounting for 54.5% of that group's events, significantly more than the intermediate (31.6%) and low-risk (20.0%) groups (P=0.01); the peak was between 05:00 and 07:00, consistent with morning toileting and residual sedative effects. Overall, falls most often occurred at the bedside (42.1%) and in the bathroom or shower room (36.8%); the high-risk group had the highest proportion of bathroom falls (48.5%), while the low-risk group had a relatively higher proportion of bedside falls (50.0%). Regarding precipitating activity, high-risk falls usually occurred during toileting or rising from bed (72.7%), whereas low-risk falls were often related to distraction while walking or slippery floors (60.0%). These patterns indicate that high-risk falls stem mainly from internal balance and mobility impairment, while low-risk falls are more often triggered by the external environment.

Independent risk factors and interactions

Multivariable logistic regression (entering variables with P<0.10 on univariate analysis and adjusting for age, sex, and disease severity) confirmed the following independent risk factors: use of sedative-hypnotics (adjusted OR=3.47, 95% CI 2.31-5.22, P<0.001), prior fall history (OR=3.12, 95% CI 2.05-4.75), high-risk Morse grade (OR=3.12, 95% CI 1.76-5.53), antipsychotics (OR=2.89), impaired activity (Barthel <60, OR=2.56), antidepressants (OR=2.45), polypharmacy with five or more drugs (OR=2.14), night-time toileting at least twice per night (OR=2.01), cognitive impairment (MMSE <24, OR=1.89), and antihypertensive use (OR=1.68). The model was statistically significant (chi-square=89.45, P<0.001) with good fit (Hosmer-Lemeshow P=0.342). In the high-risk subgroup, patients with both polypharmacy and frequent night-time toileting had a fall rate of 18.5 per 1,000 bed-days, significantly higher than those with only one factor (11.2 and 9.8 per 1,000 bed-days; interaction P=0.021); the interaction weakened in the intermediate group (P=0.089) and was absent in the low-risk group (P=0.452). This implies that interventions for high-risk patients must address both medication review and environmental modification. Teams can use QSevidence to compare these effect sizes with published studies of psychotropic drugs and fall risk, verify consistency, and produce a citable evidence summary.

Secondary outcomes: length of stay, fear of falling, and activity

The intervention group stayed 12.4±5.8 days, shorter than the control group's 14.1±6.5 days (P=0.004). Baseline FES-I scores were similar (32.5±8.1 versus 33.1±8.4, P=0.512); at discharge, FES-I was lower in the intervention group (28.1±7.6 versus 31.2±8.0, P<0.001), with a larger reduction (-4.4±5.2 versus -1.9±4.8, P<0.001). Barthel scores were similar at baseline (62.3±18.5 versus 61.9±19.1, P=0.834); at discharge the intervention group scored higher (70.1±17.2 versus 65.8±18.6, P=0.015), with greater improvement (+7.8±10.4 versus +3.9±9.7, P<0.001). Effective prevention did not restrict activity; by reducing falls and their injuries it preserved continuity of rehabilitation and promoted functional recovery.

Discussion: Drug-Environment Interactions, Targeted Care, and Resource Focus

This study shows that Morse-based tiered care can reduce falls in older inpatients by about 44%. The effect rests on two mechanisms. Targeted intervention matches care to each patient's dominant risks: for high-risk patients it goes beyond environmental modification to emphasize medication review and intensified supervision, consistent with guideline-endorsed multifactorial intervention. Resource focus concentrates limited nursing time, anti-fall equipment, and supervision on high-risk patients, avoiding over-treatment of low-risk patients and, from a health-economic perspective, is expected to reduce fractures and their high costs. The study also reveals a multiplicative interaction between psychotropic drugs and night-time toileting as key risk points: sedative-hypnotics increase risk through drowsiness, dizziness, orthostatic hypotension, and delayed reaction, while night-time toileting adds poor lighting, privacy concerns, and sudden postural change, so that drug, physiologic, and environmental factors converge in the middle of the night and trigger falls in high-risk patients.

Three practical measures follow. First, use the Morse or Hendrich II scale as a routine assessment within 24 hours of admission and trigger the corresponding intervention package automatically by score. Second, embed risk assessment in the electronic health record to enable automatic scoring and alerting, reducing manual work, and appoint fall-prevention specialist nurses for case management and quality checks. Third, strengthen multidisciplinary collaboration with closed-loop quality control: pharmacists lead medication review in high-risk patients (targeting antipsychotics, benzodiazepines, and diuretics), rehabilitation therapists design individualized training, environmental safety teams regularly inspect lighting and slip resistance, and a quarterly cycle of root-cause analysis, intervention revision, and re-evaluation is maintained. Within this system, the QSevidence medical AI tool provides clinical decision support: AI guideline retrieval checks the recommendation sources for grading standards and each tier of intervention, literature evidence appraisal tracks the latest evidence on psychotropic drugs and night-time falls and evaluates the strength of medication-review protocols, and structured evidence generation organizes risk-factor odds ratios, event patterns, and intervention packages into a reviewable evidence table that supports multidisciplinary discussion, root-cause analysis, and quality review, moving fall prevention from experience-driven to evidence-driven care.

Limitations and Future Directions

This single-center study limits generalizability to other regions and facility levels; the nature of the intervention precluded blinding of patients and staff, introducing possible performance and detection bias; the actual sample (365) was below the prespecified target (2,748), limiting the power of subgroup analyses; and follow-up ended at discharge, so post-discharge falls, function, and readmission were not captured. Future directions include multicenter cluster randomized controlled trials with stratified randomization and independent outcome assessment; propensity-score matching or inverse probability weighting to correct non-randomized bias; machine-learning integration into risk-prediction models for more accurate, dynamic stratification; and follow-up of 3-6 months after discharge with patient-reported outcomes such as fear of falling, functional independence, and quality of life, plus cost-effectiveness analysis. In such prospective research, protocol registration, demonstration of baseline comparability, and outcome definition can be systematically prepared and reported with the literature appraisal and structured evidence generation capabilities of QSevidence, improving reproducibility and clinical translation.

References

  1. World Health Organization. WHO Global Report on Falls Prevention in Older Age. Geneva: WHO; 2007.
  2. Morse JM. Preventing Patient Falls: Establishing a Fall Intervention Program. 2nd ed. New York, NY: Springer Publishing Company; 2008.
  3. Panel on Prevention of Falls in Older Persons, American Geriatrics Society and British Geriatrics Society. Summary of the Updated American Geriatrics Society/British Geriatrics Society clinical practice guideline for prevention of falls in older persons. J Am Geriatr Soc. 2011;59(1):148-157.
  4. National Institute for Health and Care Excellence. Falls in older people: assessing risk and prevention. NICE guideline [CG161]. London: NICE; 2013.
  5. Cameron ID, Dyer SM, Panagoda CE, et al. Interventions for preventing falls in older people in care facilities and hospitals. Cochrane Database Syst Rev. 2018;(9):CD005465.
  6. Yardley L, Beyer N, Hauer K, et al. Development and initial validation of the Falls Efficacy Scale-International (FES-I). Age Ageing. 2005;34(6):614-619.
  7. Mahoney FI, Barthel DW. Functional evaluation: the Barthel index. Md State Med J. 1965;14:61-65.
  8. Hendrich A, Nyhuis A, Kippenbrock T, Soja ME. Hospital falls: development of a predictive model for clinical practice. Appl Nurs Res. 1995;8(3):129-139.
  9. QSevidence official website (qsevidence.com): AI guideline retrieval, literature evidence appraisal, and structured evidence generation tool.

Medical Disclaimer

This article is based on a single-center prospective cohort study and publicly available literature. It is provided for medical education and evidence-based nursing research reference only and does not constitute individualized diagnosis or treatment advice. Fall risk assessment, grading standard selection, and preventive measures must be carried out by qualified clinicians according to each patient's condition and current guidelines.