Emergency Medicine · Medical Literature
Emergency department-initiated interventions to reduce risk of recurrent falls in older adults: a systematic review and meta-analysis.
Abstract
Older adults presenting to the emergency department (ED) following a fall are at high risk for recurrent falls, injuries and increased healthcare utilisation. This systematic review and meta-analysis evaluated the effectiveness of interventions initiated upon ED presentation in reducing falls (primary outcome) and fall-related outcomes (secondary outcomes) among older adults. A comprehensive search was performed in Ovid Medline, Embase, CINAHL, PEDro, Web of Science, and Scopus up to June 2025, following PRISMA guidelines. This yielded 9624 references, of which 4811 records remained after deduplication and were screened for eligibility. Meta-analyses and descriptive methods, including vote counting, were applied where appropriate. Sensitivity and subgroup analyses were conducted. Thirty articles were included. Interventions were associated with a significant reduction in fall incidence (rate ratio 0.69; 95% CI 0.54-0.88; I2 = 95%), supported by vote counting. While reductions were also observed in the proportion of fallers (odds ratio 0.87; 95% CI 0.71-1.07; I2 = 61%), fall-related and all-cause ED revisits, hospital admissions, recurrent falls and mortality, these trends did not reach statistical significance. Sensitivity analysis excluding studies with modified usual care showed a significant reduction in fall-related ED revisits and people sustaining an injurious fall. Secondary outcomes including quality of life, fear of falling and physical functioning generally showed trends favouring the intervention group. Considerable heterogeneity and variation in intervention characteristics were noted across studies. Interventions for older adults presenting to the ED after a fall suggest a favourable effect on fall rate. Non-significant trends favouring the intervention group were observed for several other outcomes, but heterogeneity and methodological limitations preclude definitive conclusions.
Clinical relevance
In older adults presenting to the emergency department following a fall, do ED-initiated interventions significantly reduce recurrent falls, fall-related outcomes, healthcare utilisation, and mortality compared with usual care? __SNAPSHOT__ {"studyDesign":"Systematic review and meta-analysis (PRISMA-compliant); combined meta-analytic and descriptive vote-counting methods","setting":"Not reported (multi-national; studies identified from Ovid Medline, Embase, CINAHL, PEDro, Web of Science, and Scopus through June 2025)","sampleSize":"30 articles included from 4,811 deduplicated records screened (9,624 total references identified); aggregate patient-level n not reported in abstract","population":"Older adults presenting to the ED following a fall; specific age thresholds and comorbidity profiles of included primary studies not reported in abstract","intervention":"Interventions initiated upon ED presentation targeting fall prevention (specific intervention types — e.g., multifactorial assessment, exercise referral, medication review — not individually enumerated in abstract)","comparator":"Comparator arms not explicitly described in abstract; assumed usual care or no intervention based on study design","primaryOutcome":"Fall incidence (rate ratio); secondary outcomes included proportion of fallers, fall-related ED revisits, all-cause ED revisits, hospital admissions, recurrent falls, and mortality","followUp":"Not reported","fundingSource":"Not reported"}
Key findings
The primary outcome of fall incidence showed a statistically significant reduction associated with ED-initiated interventions: rate ratio (RR) 0.69 (95% CI 0.54–0.88), corresponding to a relative reduction of approximately 31% in fall incidence. This result was supported by vote counting. However, between-study heterogeneity was very high (I² = 95%), substantially limiting confidence in the pooled estimate. The secondary outcome of proportion of fallers (odds ratio 0.87; 95% CI 0.71–1.07; I² = 61%) did not reach statistical significance, indicating that while there was a trend toward fewer individuals experiencing at least one fall, this could not be distinguished from chance. The abstract further reports non-significant reductions in fall-related and all-cause ED revisits, hospital admissions, recurrent falls, and mortality; however, point estimates, confidence intervals, and p-values for these secondary outcomes are not reported in the available abstract. NNT cannot be calculated from the data provided. No clinically important adverse events were reported in the abstract.
Practice implications
This meta-analysis suggests that ED-initiated interventions may reduce the rate of falls in older adults who present following a fall, but several important caveats apply. First, the primary outcome (fall incidence, RR 0.69) achieved statistical significance, but the accompanying I² of 95% represents extreme heterogeneity, meaning the pooled estimate integrates studies with markedly different interventions, populations, follow-up durations, and control conditions. The true effect in any specific clinical context could be substantially larger or smaller than the pooled RR. Second, the proportion of fallers — arguably the more patient-centred binary outcome — showed only a non-significant trend (OR 0.87, 95% CI crossing 1.0), weakening confidence that these interventions consistently prevent individuals from re-falling. Third, none of the secondary outcomes (ED revisits, hospitalisations, recurrent falls, mortality) achieved significance per the abstract, meaning clinically important downstream benefits on healthcare utilisation and patient safety remain unestablished. The findings are not directly applicable to individual patients: it is unclear which intervention type, intensity, or patient subgroup drives any observed benefit. Patients with cognitive impairment, severe frailty, or polypharmacy — highly prevalent among ED fall presenters — may not be adequately represented. The results are hypothesis-generating and support continued investment in ED-based fall-prevention infrastructure, but they are insufficient to mandate a specific clinical protocol. IMPORTANT NOTE: This summary is based solely on the provided abstract; full-text data including subgroup analyses, individual study characteristics, intervention taxonomy, and risk-of-bias assessments are unavailable and may materially alter interpretation. __LIMITATIONS__ Extreme statistical heterogeneity (I² = 95% for the primary outcome) renders the pooled rate ratio difficult to interpret clinically; it likely reflects pooling of fundamentally disparate interventions rather than a coherent treatment effect.,The primary pooled effect (RR 0.69) uses fall incidence (a rate), while the non-significant secondary result uses proportion of fallers (a binary outcome); these measure different constructs, and their discordance is unexplained in the abstract.,The secondary outcomes — including the arguably most clinically meaningful endpoints of ED revisits, hospitalisation, and mortality — did not reach statistical significance; quantitative estimates for these are not reported in the abstract.,Nature, dose, and fidelity of the interventions across 30 included articles are not described in the abstract, precluding determination of which specific ED-initiated intervention is effective.,No follow-up duration is reported, making it impossible to assess durability of any observed benefit.,Aggregate patient-level sample size, risk-of-bias distribution across included studies, and blinding status are not reported in the abstract.,Publication bias was not addressed in the abstract; positive trials may be over-represented.,Generalisability is uncertain: patient age thresholds, frailty status, cognitive function, and healthcare system context of included studies are unreported.,Funding source for the systematic review is not reported, precluding assessment of potential conflict-of-interest influence.,Only an abstract was available for this summary; full-text review may reveal additional limitations or nuances not captured here. __BOTTOM_LINE_CATEGORY__ promising_but_preliminary __BOTTOM_LINE__ ED-initiated interventions were associated with a statistically significant reduction in fall incidence (RR 0.69; 95% CI 0.54–0.88), but extreme heterogeneity (I² = 95%) and the failure of secondary outcomes — including proportion of fallers, ED revisits, hospitalisations, and mortality — to reach statistical significance substantially temper this finding. Clinicians should view these results as supportive evidence for pursuing structured ED-based fall-prevention programs rather than as validation of any specific protocol, pending higher-certainty evidence from adequately powered trials with standardised interventions and clearly defined populations.