Why Fracture Patients Fall Through the Cracks: Depression, Fear of Falling and the Missing Data in Bone-Care Pathways

The bone is repaired, but the underlying osteoporosis often goes untreated, setting the stage for the next fracture. Fracture liaison services, or FLSs, were invented to close that gap: coordinated teams that systematically identify fracture patients, assess their bone health, recommend treatment and follow up. The model works, on paper.

Yet a striking number of eligible patients still vanish somewhere between identification, assessment, treatment recommendation, medication initiation, rehabilitation and follow-up. A new policy review published in Frontiers in Public Health asks an uncomfortable question: could depression and fear of falling be part of the reason, and should mental health be built into fracture services at all? The review, led by Fujin Liao and colleagues at Shenzhen Pingle Orthopedic Hospital and Shenzhen University, is emphatic that the answer is not a simple yes.

Drawing on structured PubMed searches that yielded 746 unique records, of which 41 sources were included in the final synthesis, the authors examined how international FLS frameworks, clinical guidelines and national audits handle the moments when patients fail to progress. Their central finding is a measurement blind spot: existing quality indicators can tell you whether a service step happened, but they cannot consistently tell you why it did not, who attributed the reason, what response was delivered, or what happened next. The psychological evidence they assembled is suggestive but carefully bounded.

A meta-analysis of 27 studies covering 11,958 older people with hip fracture estimated a pooled depression prevalence of 23 percent, though estimates varied widely by setting and measurement method. In a Danish population-based cohort of 57,887 hip-fracture patients aged 65 or older, the one-year cumulative incidence of anti-osteoporosis medication initiation was 14.8 percent among patients with a recorded mood disorder versus 20.8 percent among those with no recorded mental disorder. Lower initiation also appeared for organic, schizophrenic, substance-use and anxiety disorders, meaning the treatment gap was not specific to depression.

A US Medicare analysis similarly linked antidepressant use to lower post-fracture treatment initiation in both women and men. But here is the catch that the authors hammer home: these are associations, not causes. Neither the Danish nor the Medicare study measured depressive symptoms, treatment beliefs, clinician recommendations and FLS participation within the same care pathway.

Administrative mental-disorder diagnoses may instead capture differences in frailty, cognition, residence, comorbidity and continuity of care. Antidepressant dispensing is particularly slippery, since it may signal current depression, past depression, anxiety, pain treatment or simply good access to healthcare. The review treats such findings as prognostic signals, potential flags identifying groups at risk, rather than established explanations for disengagement.

Fear of falling tells a parallel story with its own subtleties. The authors stress that concern about falling is conceptually distinct from objectively assessed falls risk: one is a clinical estimate of future falls based on history, physical performance, medication, cognition, vision and environment; the other is the patient's own appraisal of falling and its consequences, including balance confidence and activity avoidance. A patient can carry substantial objective risk without marked worry, or be profoundly activity-restricted by fear despite modest measured risk.

These presentations demand different responses, and the current FLS falls-assessment indicator cannot distinguish between them. A systematic review of 36 studies involving 5,099 participants found reported fear-of-falling prevalence after hip fracture ranged from 22.5 percent to 100 percent, a spread reflecting wildly different instruments, thresholds and timing rather than a single credible figure. The intervention evidence adds a sobering twist.

A systematic review of eight randomized trials with 1,146 participants found that exercise rehabilitation after hip fracture showed a pooled end-of-intervention effect on depressive symptoms favoring rehabilitation, but statistically imprecise, with a Hedges' g of minus 0.43 and a 95 percent confidence interval from minus 0.87 to 0.01. A cognitive-behavioral trial called Step by Step improved falls-related self-efficacy, with an effect size of minus 0.42 on the Short Falls Efficacy Scale-International, yet did not increase objectively measured daily walking duration. And a quasi-experimental study combining a patient decision aid with motivational interviewing improved patients' reported involvement in treatment decisions but did not significantly improve one-year medication persistence, 80.4 percent versus 76.7 percent under usual care.

Crucially, no trial identified in the review used completion of an FLS engagement node as its primary outcome. Improved questionnaire scores, in other words, cannot be assumed to mean more patients taking medication or attending rehabilitation. To make the policy gap concrete, the authors use the England and Wales Fracture Liaison Service Database as a worked case.

The 2026 annual report covers care delivered during 2024 to more than 83,500 patient records for people aged 50 or older across 77 participating services, with audit coverage estimated at 63 percent of acute hospitals treating fragility fractures. Reported national results included 69 percent for bone-health assessment within 90 days, 41 percent for DXA within 90 days, 57 percent for treatment recommendation, 9 percent for strength and balance commencement and 30 percent for treatment recorded at 12 months. But these percentages are not a single cohort funnel: their denominators, eligibility rules and time windows differ, and follow-up indicators are conditioned on prior treatment recommendation.

A patient lost before recommendation never enters later denominators; a patient recorded as declining may be excluded; a non-contactable patient may remain in the 12-month denominator. The audit can locate weak processes but cannot reconstruct the probability that the same patient passes through every transition. The authors' remedy is deliberately modest and staged.

First, cross-node reason attribution: whenever an eligible FLS step is not completed, the service records one primary reason and any secondary contributors, keeps patient, caregiver, clinician and administrative accounts source-labeled, and allows unknown as a valid answer. The proposed taxonomy distinguishes unsuccessful contact, informed refusal, unresolved treatment uncertainty, clinical ineligibility, cognitive or communication support needs, patient-reported psychological barriers, practical access problems, service failure and unknown or mixed causes. Psychological attribution requires patient or caregiver report or documented clinical assessment linking the factor to the specific node; a questionnaire score or a missed appointment alone is not enough.

Second, targeted case finding only where response capacity exists, using brief two-stage instruments such as the PHQ-2 followed by the PHQ-9 for depressive symptoms, or the Falls Efficacy Scale-International for activity-limiting concern, always tied to a named response and responsible actor. Third, multicentre evaluation before any universal stepped pathway is considered. That restraint reflects a sharp ethical point running through the review: a considered refusal after an accessible, balanced discussion is a legitimate outcome, not a failure.

A missed appointment should not be coded as depression when contact was inaccessible or interpretation unavailable. Combining informed choice with remediable barriers under the umbrella of non-adherence prevents services from locating, and therefore correcting, the actual loss. Equity concerns compound this: questionnaire completion and follow-up tend to be lower among people with cognitive impairment, limited literacy, sensory loss, language barriers or digital exclusion, and removing them from denominators can make an intervention look feasible by selectively excluding those who most need adaptation.

The review insists that evaluation report reach, missingness and target-node completion across equity strata, with prespecified stopping rules if core fracture-prevention performance deteriorates. The bottom line is a reframing of what mental-health integration in fracture care should mean. The authors explicitly do not propose turning FLSs into general mental-health services or adding universal depression screening to international accreditation.

Instead, they argue that mental-health integration should be treated as a testable quality-improvement strategy: make pathway loss attributable first, add targeted measurement only where a response can be delivered, and prove the combined pathway across multiple services before setting standards. Success, they conclude, should be judged not by the number of questionnaires completed, but by whether the intervention reaches patients equitably, produces an appropriate response, improves the targeted engagement node and preserves the core work of secondary fracture prevention. For a field where nearly a third of patients lack documented treatment at twelve months, that measured, auditable approach may be the most viral idea of all: before screening everyone, first learn why the system loses people in the first place.

Subject of Research: Integrating depression and concerns about falling into fracture liaison service policy through attribution-first measurement of patient pathway loss

Article Title: Integrating mental health into fracture liaison services: an attribution-first policy approach to depression, concerns about falling and pathway loss

Article References: Liao, F., Zhou, F., Li, X., Wang, Z., & Wang, Q. (2026). Integrating mental health into fracture liaison services: an attribution-first policy approach to depression, concerns about falling and pathway loss. Frontiers in Public Health, 14, Article 1989494. https://doi.org/10.3389/fpubh.2026.1989494

Image Credits: AI Generated

DOI: 10.3389/fpubh.2026.1989494

Keywords: fracture liaison services, osteoporosis, secondary fracture prevention, depression, fear of falling, patient engagement, health policy, quality indicators, implementation science, hip fracture, mental health integration, clinical audit