Secondary Logo

Journal Logo

Featured Articles

The Association of Frailty With Adverse Outcomes After Multisystem Trauma: A Systematic Review and Meta-analysis

Poulton, Alexander MD*; Shaw, Julia F. BSc; Nguyen, Frederic MD*; Wong, Camilla MD, FRCPC; Lampron, Jacinthe MD, FRCSC§,‖; Tran, Alexandre MD, MSc†,§; Lalu, Manoj M. MD, PhD, FRCPC*,†,¶,#**; McIsaac, Daniel I. MD, MPH, FRCPC*,†,¶,#

Author Information
doi: 10.1213/ANE.0000000000004687



  • Question: What is the association of preinjury frailty with mortality, morbidity, resource use, and other health and experiential outcomes after multisystem trauma?
  • Findings: In this systematic review and meta-analysis of 16 studies (5198 participants), frailty was associated with a significant increase in mortality, in-hospital complications, and adverse discharge disposition.
  • Meaning: Frailty represents an important prognostic factor in multisystem trauma; future research is needed to optimize trauma systems of care for the unique needs of older people with trauma and to standardize assessment.

Western populations are aging rapidly; by 2050, the number of people ≥65 years of age in the United States is expected to double relative to the present.1 The aging population directly impacts trauma care because older patients are presenting with multisystem trauma at increasing rates.2 In 2016, adults >65 years of age accounted for 31% of trauma incidents and had the highest fatality rates.3 Frailty, a multidimensional condition related to accumulation of age- and disease-related deficits leading to vulnerability to stressors,4,5 increases in prevalence with age,6 and is a robust predictor of adverse outcomes in older people experiencing acute and chronic conditions.7–9

As the proportion of traumatic injuries accounted for by older people continues to increase, understanding the relationship between preinjury frailty and trauma outcomes will be imperative. This information could help to inform risk stratification at the time of trauma team activation, discussions around goals of care with patients and their families, as well as geriatric-specific modifications to trauma systems of care. To date, while frailty instruments and their application in trauma have been reviewed,10,11 frailty-related trauma outcomes have not been systematically and quantitatively synthesized. Furthermore, because of the geriatric-specific nature of many trauma presentations in older people (such as multisystem injuries with relatively low energy mechanisms),2,12 synthesis of individual study results must consider potential sources of confounding that bias the association between preinjury frailty and subsequent outcomes.

To address this key knowledge gap, our objective was to conduct a systematic review and meta-analysis to estimate the association of pretrauma frailty with mortality (primary outcome), as well as secondary outcomes informed by the Institute for Healthcare Improvement’s Triple Aim (health, resource use, and experience outcomes).13–15


Following protocol registration (International Prospective Registry of Systematic Reviews [CRD42018104116]), we conducted a systematic review and meta-analysis of observational studies following recommendations of the Meta-analysis Of Observational Studies in Epidemiology group.16 We also followed the recommendations of the Cochrane Collaboration in designing and conducting our review,17 along with recent recommendations for systematic reviews of prognostic studies.18 The results are reported in keeping with the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) Statement.19

Search Strategy

We developed a comprehensive search strategy (Supplemental Digital Content, eTable 1, in consultation with an information specialist; the strategy was finalized after undergoing peer review.20 Keywords for frailty and trauma were combined and the strategy was applied to MEDLINE, EMBASE, and the Comprehensive Index to Nursing and Allied Health Literature (CINAHL) databases from inception to May 22, 2019. We did not include abstracts or sources of grey literature as methodological descriptions would have been insufficient to assess frailty definitions, study quality, and risk of bias.21 The reference lists of related systematic reviews as well as included articles were searched by hand to identify other studies that may have been missed by our initial search. No language restrictions were applied.


Our primary outcome was all-cause mortality at any point after the initial trauma (as mortality has been identified as an outcome of primary importance for older people).22 Secondary outcomes were informed by the Institute for Healthcare Improvement’s Triple Aim of health outcomes, experience, and costs.15 Specifically, under the health outcomes domain, we evaluated complications (as reported in each study), adverse discharge disposition (ie, death or discharge to a nonhome location or higher level of care than before the traumatic injury), falls, and postinjury frailty. Post hoc delirium was also considered. Under the costs domain, we evaluated length of stay (LoS) and health care costs, and under experience, we evaluated any measure of patient or caregiver experience/satisfaction.

Inclusion and Exclusion Criteria

Eligible studies were included if they met the following criteria: (1) acute multisystem trauma (ie, an acute injury involving more than one body system requiring trauma team activation); (2) participants ≥18 years of age; (3) pretrauma frailty provided as an exposure using an explicitly described frailty instrument; (4) reported relevant outcomes and relevant quantitative outcome data (ie, event rates or measures of association estimating the relationship between frailty and outcomes). Studies were excluded if they (1) lacked a comparator group (eg, case reports or case series); (2) reported isolated injuries (ie, isolated fragility fractures, burns, head injuries); (3) reported mixed trauma and nontrauma populations for which the trauma-specific population was <50% of the sample or trauma-specific data could not be extracted.

Selection of Included Studies and Data Extraction

First, duplicate and independent review of titles and abstracts was performed. Studies reviewed as “yes” or “unsure” were advanced to full-text review; agreement between both reviewers was required to exclude studies. The full-text review was also completed in duplicate by independent reviewers. Any disagreements were resolved by consensus in discussion with the senior author (D.I.M.). Studies that were included after full-text review then underwent data extraction using a form specifically designed for this study; the form was piloted by 2 reviewers before full implementation. Data were extracted by 1 reviewer and independently checked for accuracy by a second reviewer and the senior author. Study authors were contacted as required to request missing or incomplete data and to clarify methods or findings. All stages of the review were completed using DistillerSR (Evidence Partners, Ottawa, ON, Canada), a web-based systematic review platform.

Statistical Analysis

All analyses were completed using Comprehensive Meta Analysis (Biostat, Englewood, NJ). Meta-analyses used random-effects models using inverse variance weighting according to the methods of DerSimonian and Laird.23 As recommended for reviews of prognostic factors,18 the primary analyses for binary outcomes pooled adjusted effect estimates that controlled for a prespecified set of confounders (injury severity and blunt versus penetrating trauma). This required pooling of odds ratios (ORs). Unadjusted analyses were also completed. For the continuous outcome LoS, we pooled means and standard deviations from each study reporting the outcome of interest; where medians and interquartile ranges were reported we transformed these values to means and standard deviations using the methods of Wan et al.24 Heterogeneity was assessed using the I2 statistic, but this value was not used to guide analytic approaches, which were prespecified as random effects models as we did not think our data would meet the assumptions required for the less conservative approach of fixed effects modeling. Hazard ratios (HRs) were handled separately from ORs. For the primary outcome (mortality), we pooled results from the most distal time period reported. A funnel plot was constructed to examine asymmetry possibly attributable to publication bias or other sources, as well as conduct of a post hoc Egger’s test.25,26 We also performed an exploratory metaregression, testing average injury severity score as an effect modifier between frailty and the odds of mortality. Finally, we separately pooled mortality results by short-term (30 days or in-hospital) versus long (beyond 30 days or in-hospital)-term follow-up to evaluate whether this led to differences in pooled ORs. Where inadequate data were available to support a meta-analysis, results were narratively analyzed. A 2-tailed, 5% significance level was used for all analyses.

Assessment of Risk of Bias

Risk of bias was analyzed using the Cochrane Collaboration’s Risk of Bias in Nonrandomized Studies-of Interventions (ROBINS-I) tool.27 Risk of bias was assessed for each study by the senior author and independently by a second team member. A score of low, moderate, high, or critical risk was assigned for each of confounding bias, selection bias, measurement bias (outcome or exposure), missing data bias, and selection bias. Disagreements were resolved through consensus. Post hoc, we explored whether excluding studies rated as high risk of bias were from meta-analyses would lead to substantively different effect estimates.


Figure 1.
Figure 1.:
PRISMA flow diagram for study selection and inclusion. CINAHL indicates Comprehensive Index to Nursing and Allied Health Literature; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-analysis.

We identified 2415 titles; 73 were advanced for full-text review (PRISMA diagram, Figure 1). After full-text review, 16 studies were included in our study; reasons for exclusion at full-text review are provided in Figure 1. All articles were published in English between 2011 and 2019.

Participants and Study Characteristics

Table 1. - Summary of Included Studies
Study (Year) Study Design Country of Origin Study Size, n Mean Age Female (%) Average ISS Injury Mechanism Frailty Measure Frail (%) Primary Outcome
Kaplan et al28 (2017) Retrospective United States 450 75.6 40.2 14.7 Mixed CT Sarcopenia/osteopenia 64.2 Mortality
Maxwell et al21 (2016) Prospective United States 188 77.0 56.0 10 Blunt VES-13 a Mortality
Joseph et al29 (2014) Prospective United States 250 77.9 30.8 15 Mixed Frailty Index 44.0 In-hospital complications
Couch et al28 (2018) Retrospective Australia 205 76.9 45.0 20.5 Mixed CT Psoas sarcopenia a Mortality
Min et al30 (2011) Prospective United States 63 78.0 33.0 14 Mixed VES-13 a Mortality
Cheung et al31 (2017) Retrospective Canada 260 76.7 47.3 17 Mixed CFS 63.5 Discharge disposition
Joseph et al32 (2017) Prospective United States 350 78.0 40.0 11 Blunt TFSI 64.0 Trauma readmission, falls
Joseph et al33 (2016) Prospective United States 368 74.8 38.6 11 Blunt TFSI 75.0 Complication-related mortality
Joseph et al34 (2014) Prospective United States 200 77.0 28.0 15 Blunt TSFI a Discharge disposition
Ebbeling et al35 (2014) Retrospective United States 180 74.0 43.0 24 Mixed CT PLVI 50.0 Morbidity, mortality
Joseph et al29 (2014) Prospective United States 100 76.5 41.0 14 Blunt Frailty Index a Discharge disposition
Curtis et al36 (2018) Retrospective United States 1403 77.6 a a Mixed CFS a Mortality, discharge disposition
Hamidi et al37 (2018) Prospective United States 267 76.9 36.4 a Mixed TSFI 60.2 Functional improvement
Lauerman et al38 (2019) Retrospective United States 489 a 40.0 a Blunt Radiographic marker composite score 28.0 Discharge disposition
Mccusker et al39 (2019) Prospective United States 325 76.0 36.0 11 Mixed TSFI 36.0 In-hospital complications, in-hospital mortality
Tipping et al40 (2019) Prospective Australia 100 69.2 19.0 22 a CFS 13.0 a
Abbreviations: CFS, Clinical Frailty Scale; CT, computed tomography; FI-Lab, Frailty Index-Lab; ISS, Injury Severity Scale; PLVI, Psoas:Lumbar Vertebra Index; TFSI, Trauma-Specific Frailty Index; VES-13, Vulnerable Elders-13 Survey.
aInformation not available.

In total, studies included 5198 participants. All included studies were cohort designs (10 prospective and 6 retrospective). Average study population age ranged from 58 to 78; 19%–56% of participants were female. The majority of studies included a mixed population exposed to either blunt or penetrating trauma (59% mixed, 35% blunt; see full description of included studies, Table 1).

Preinjury Frailty and Outcomes

From 10 studies that allowed calculation of frailty prevalence, 1581 of 3039 (52%) participants had some degree of frailty, which was assigned using 8 unique frailty instruments (Psoas muscle cross-sectional area [n = 1],41 L3 sarcopenia/osteopenia [n = 1],28 Vulnerable Elders Survey [n = 2],21,30 Clinical Frailty Scale [n = 3],31,36,40 Psoas:lumbar vertebral index [n = 1],35 Frailty Index [n = 2],29,42 Radiographic marker composite score [n = 1],38 and Trauma-Specific Frailty Index [n = 5]32–34,37,39). Timing of administration of frailty assessments was not clearly described in most studies; frailty instruments are described in (Supplemental Digital Content, eTable 2,


Table 2. - Summary of Primary Outcome—Mortality
Study (Year) Frailty Measure Distal Time Point Reported Frail, n Frail Dead, n Without Frailty, n Without Frailty Dead, n Adjustment Variables (If Applicable) Adjusted OR (95% CI) P
Joseph et al42 (2014) Frailty Index Hospital discharge 110 5 140 0
Cheung et al31 (2017) CFS Hospital discharge 158 7 95 4
Ebbeling et al35 (2014) CT PLVI Hospital discharge 90 14 90 9 Age, comorbidity, h-AIS, ISS 1.200 (0.441–3.266) .721
Joseph et al32 (2017) TFSI 6 mo 224 19 126 4 Age, gender, ISS, h-AIS, injury mechanism, in-hospital complications 1.100 (1.039–1.164) .001
Joseph et al33 (2016) TFSI Hospital discharge 275 30 93 6
Mccusker et al39 (2019) TSFI Hospital discharge 117 15 208 10 Demographics, admission vitals, injury parameters, comorbidities, sarcopenia 1.720 (1.510–1.960) <.001
Tipping et al40 (2019) CFS Hospital discharge 13 4 87 8
Maxwell et al20 (2016) VES-13 12 mo 94 Age, ISS, comorbidity index, AD8 1.330 (1.074–1.647) .009
Min et al30 (2011) VES-13 Hospital discharge ISS, sex, CCI 1.070 (0.661–1.372) .783
Curtis et al36 (2018) CFS Hospital discharge Age, GCS, ISS 1.560 (1.408–1.728 <.001
Kaplan et al28 (2017) CT Sarcopenia/osteopenia 12 mo 289 32 161 1
Abbreviations: AD8, Ascertain Dementia 8 Score; AIS, Abbreviated Injury Scale; CI, confidence interval; CCI, Charleston Comorbidity Index; CFS, Clinical Frailty Scale; CT, computed tomography; GCS, Glasgow Coma Scale; h-AIS, Head Abbreviated Injury Scale; ISS, Injury Severity Scale; OR, odds ratio; PLVI, Psoas:Lumbar Vertebra Index; TFSI, Trauma-Specific Frailty Index; VES-13, Vulnerable Elders-13 Survey.

Figure 2.
Figure 2.:
Forest plot of adjusted mortality data. CI indicates confidence interval.

Overall, from the 8 studies that allowed calculation of event rates, 126 of 1276 (9.9%) people with frailty died, compared to 42 of 1000 (4.2%) people without frailty. Mortality ascertainment windows ranged from 30 days or in-hospital31,33,35,39,40,42 to 6 months33 or 1 year.28 Four studies28,30,32,35 provided unadjusted mortality data that allowed for meta-analysis. In these studies, the odds of mortality were higher with frailty, but not significantly so (unadjusted OR, 2.75; 95% CI, 0.91–8.30, I2 = 75.9%). Seven studies21,28,30,32,35,36,39 provided mortality data adjusted for our minimum set of confounders. In the 6 studies21,30,32,35,36,39 that reported adjusted ORs for mortality (and could therefore be combined in meta-analysis), frailty was associated with a significant increase in the odds of mortality (adjusted OR, 1.53; 95% CI, 1.37–1.71, I2 = 29.9%; Table 2, forest plot, Figure 2). A single study reported HRs, which could not be used in the meta-analysis; within this study, low, moderate, and severe frailty were all associated with decreased survival compared to no frailty (adjusted HR versus no frailty: 14.4 [95% CI, 1.6–128.7]-low; 17.9 [95% CI, 2.4–134.4]-moderate; 26.7 [95% CI, 3.4–206.7]-high).28 The funnel plot (Supplemental Digital Content, eFigure 1, demonstrated asymmetry; however, possibly missing studies were in the quadrant of the plot where smaller studies with large effect sizes would be expected; this suggests that publication bias was not the cause of asymmetry.25 Furthermore, the 1-tailed P value from Egger’s test for asymmetry was 0.089, which did not support significant asymmetry. In a post hoc exploratory metaregression, injury severity score did not appear to be an effect modifier between frailty and mortality (P = .544). The second post hoc analysis pooled adjusted ORs from short-term follow-up (1.58, 1.41–1.79, P < .0001, I2 = 32.0)30,35,36,39 and long-term follow-up (6–12 months) and did not (1.33, 1.07–1.64, P = .009, I2 = 0)21,32 demonstrate substantial differences in effect estimates.

In-Hospital Complications

Figure 3.
Figure 3.:
Forest plot of adjusted complications data (A) and adjusted discharge data (B). CI indicates confidence interval.

Overall, 8 studies reported on in-hospital complications.28,30–33,35,39,42 Complication definitions varied between studies, but primarily focused on organ-level complications (eg, respiratory, cardiac, hematologic, infectious, renal; see Supplemental Digital Content, eTable 3,, for full description). Of 1263 people, 435 (34.4%) people with frailty had complications, compared to 202 of 913 (22.1%) of people without frailty. Five studies28,30,33,35,42 provided unadjusted data appropriate for meta-analysis, which demonstrated a significant pooled association between frailty and complications (unadjusted OR, 2.24; 95% CI, 1.33–3.77, I2 = 88.4%). Seven studies28,30,31,33,35,39,42 provided in-hospital complication data adjusted for our minimum set of confounders. These data demonstrated increased in-hospital complications among those with frailty after adjustment (adjusted OR, 2.32; 95% CI, 1.72–3.15, I2 = 55.6; Supplemental Digital Content, eTable 4,, forest plot, Figure 3A). This finding was almost unchanged with exclusion of one high risk of bias study (adjusted OR, 2.29; 95% CI, 1.64–3.19).30

Discharge Location

Twelve studies reported on discharge location.28–34,36,38–40,42 Adverse discharge disposition was classified as long-term skilled nursing facility, hospice care, or death (ie, studies classified death with nonhome discharge in a composite adverse discharge outcome). Overall, 512 of 1323 (38.7%) people with frailty had adverse discharge disposition, compared to 298 of 1262 (23.6%) of people without frailty. Five studies28–30,34,42 provided unadjusted data to the meta-analysis, and frailty was significantly associated with adverse discharge in this analysis (unadjusted OR, 1.57; 95% CI, 1.18–2.10, I2 = 77.2%). After controlling for a minimal set of confounders, 8 studies28–31,34,36,38,39,42 contributed to our adjusted meta-analysis. In these studies, the odds of adverse discharge destination was increased among those with frailty (adjusted OR, 1.78; 95% CI, 1.29–2.45, I2 = 95.8; Supplemental Digital Content, eTable 5,, forest plot, Figure 3B).

Length of Stay

Six studies reported on LoS.28,32,33,35,39,42 Average LoS ranged from 332,39 to 1635 days among patients without frailty, compared to 439 to 1735 days for people with frailty. Only unadjusted data were available for meta-analysis and demonstrated an increase in LoS when frailty was present (standardized mean difference 0.23, 95% CI, 0.12–0.34, I2 = 18.2; Supplemental Digital Content, eTable 6, eFigure 2,

Other Prespecified Study Outcomes

Among other outcomes included in the Institute for Healthcare Improvement’s Triple Aim framework, we identified studies that reported on health and cost domains; however, no experience outcomes were reported.

In the health outcome domain, studies additionally reported on postinjury frailty and falls; delirium data were not available. Maxwell et al21 found that among individuals with preinjury frailty, most had an increase in frailty 1 year after their injury. Joseph et al32 found an association between frailty and increased incidence of postinjury falls (adjusted OR, 1.6; 95% CI, 1.1–2.5).32 In the costs domain, neither Kaplan et al28 nor Min et al30 found a significant association between frailty and hospital costs or charges (respectively) during the index hospital admission.

Risk of Bias

Figure 4.
Figure 4.:
Risk of bias assessments.

The results of the risk of bias analysis for included studies are presented categorically in Figure 4. There was 77% agreement between raters across all studies and domains; 99% of ratings were in agreement within 1 level. Three studies (19%) were at low risk of bias, 10 (63%) were at moderate risk, 2 (12%) were at high risk, and 1 (6%) was at unclear risk of bias. The most common source of bias across studies was attributable to selection of participation.


In this systematic review and meta-analysis of the association between frailty and outcomes after multisystem trauma, we found that the presence of frailty was significantly associated with posttrauma mortality, as well as other adverse outcomes, including in-hospital complications, adverse discharge location, and LoS. Importantly, associations with mortality, complications, and adverse discharge locations were all significantly worse even after adjustment for confounders. Given the rapid aging of our population and the continued important role of trauma as a source of morbidity and mortality,43 these data support the role of frailty as a useful guide for risk stratification and prognosis in older trauma patients. Future efforts to build on existing frailty-tailored trauma interventions are also likely warranted.44

As frailty results in vulnerability to stressors, it is not surprising that the presence of preinjury frailty was associated with adverse outcomes after multisystem trauma. Similar associations have been described in other areas of acute care medicine, including surgery7,45 and critical care,46 for which frailty is typically associated with a >2-fold increase in mortality risk. While all available data are necessarily observational, limiting causal inference, it is highly plausible that the >1.5-fold increase in the odds of post-trauma mortality associated with frailty represents a causal association. Frailty, which results in decreased reserves, would be expected to lead to poor tolerance of the physiological and psychosocial stressors inherent to multisystem trauma. Therefore, frailty should be considered as an important part of a secondary trauma assessment for older patients as it provides important prognostic information to inform discussions with them and their families, as well as care planning.

The finding that older people with frailty also experienced greater odds of having a complication or an adverse discharge disposition is also in keeping with the larger literature associating frailty with outcomes in other clinical areas.46,47 People with frailty had more than a 2-fold increase in their odds of experiencing a serious complication and a 1.8-fold increase in their odds of not being discharged home. Knowledge of these risks is important for clinicians, patients, and those designing health systems as they plan for the predicted increase in geriatric trauma patients. Consideration of strategies to improve the care of the individual frail trauma patient, raise awareness of the associated risks, and optimize the health care system for patients with frailty is crucial. For example, enhanced geriatric processes of care emphasizing early rehabilitation could potentially decrease the risk of both complications and loss of function that could lead to a nonhome discharge. Geriatric-focused trauma wards,48 and integrated geriatric medicine consultations for trauma patients,49,50 have shown early signals of possible benefit in specific domains (such as reducing polypharmacy/high-risk medication use). Early discharge planning, which appears to be beneficial in geriatric surgical patients,51,52 might also prove beneficial for older trauma patients with frailty. These traumatic events may in fact be the first significant presentation to medical attention, necessitating assessment of frailty and focused management to try to improve outcomes and limit repeat presentations.53 Furthermore, although limited by a lack of adjusted data suitable for meta-analysis, knowledge that the presence of frailty is associated with a 20% increase in unadjusted LoS will be important for health system planners and administrators.

While it is important to consider the associations that we have identified and quantified between frailty and adverse commonly measured outcomes, it is also important to recognize the limitations in the current literature regarding other outcomes that are also of key importance to patients, their families, and clinicians. For example, our review found only 1 study reporting on postinjury frailty and functional recovery and no studies describing the experience of patients (such as issues around pain control, satisfaction, or emotional health) or their caregivers (such as burden and burnout). In keeping with previous reviews,11 we suggest that future studies are needed to evaluate and quantify the impact that frailty may have on patient- and family-centered outcomes. Resource use outcomes (eg, costs) other than LoS were also poorly described. Furthermore, across 16 studies, 8 unique frailty instruments were used, similar to previous findings described in the trauma literature.10,11 Many measures were focused on the radiologic measures of osteopenia or sarcopenia, which some consider to be distinct from, but related to frailty.54 In addition to being distinct from multidimensional frailty, these techniques require specialized radiologic techniques to quantify, which could limit application. Among multidimensional instruments, the Clinical Frailty Scale can be rapidly administered, including by proxy, making it promising for clinical practice.55,56 Full frailty indices can be difficult to routinely operationalize in practice due to the need for numerous and detailed variables, although more succinct trauma-specific indices show promise. Together, these issues clearly highlight the need for studies that provide head to head comparisons between leading frailty instruments within geriatric trauma populations to gain an understanding of what instruments are most accurate for identifying high-risk patients. These studies should also consider other aspects that will influence uptake, such as acceptability and feasibility in clinical practice.

This study should be appraised in consideration of its strengths and limitations. First, our study is grounded in a thorough search that was more extensive than previous reviews (that were limited to MEDLINE and EMBASE)11 and was performed according to a preregistered protocol and in keeping with best practice methodologies for systematic reviews and meta-analyses of observational studies.16,17 These processes should minimize bias and ensure that we have included all relevant published studies. However, due to limitations in our ability to appraise bias and quality in conference abstracts and the grey literature, these sources were not included, which could limit the generalizability and scope of our findings. There are also aspects of our review that are directly impacted by the quality of included studies. Overall, for a systematic review of observational studies, it is encouraging that risk of bias was generally low to moderate. Our review pooled outcomes that in some cases differed in terms of the outcome ascertainment timeframe. In the case of mortality, patients with frailty may be at increased risk of mortality at baseline, with prolonged duration of follow-up capturing this effect; however, the relative effect of frailty on long- and short-term did not appear to differ substantially. Also, our pooled outcomes varied in underlying definitions in some cases (eg, complications). Our prespecified analytic approach was to use random-effects meta-analysis based on content knowledge. Some authors may use a rule of thumb in which an I2 < 50% would suggest use of a fixed-effects model.57 This rule-of-thumb approach would have suggested fixed-effects analysis for mortality; therefore, our random-effects pooled OR may have unnecessarily wide 95% CIs. Our random-effects models for complications and discharge were consistent with rules of thumb. Furthermore, LoS data are typically skewed; however, meta-analysis assumes a normal distribution. This could lead to systemic error in estimation of the pooled LoS effect size. The heterogeneity of the frailty instruments used must also be considered. Eight different instruments were used and reflected a variety of approaches to frailty assessment. In particular, radiologic approaches to frailty assessment identify issues of sarcopenia or osteopenia. There remains considerable debate among geriatricians regarding the precise relationship between sarcopenia and true multidimensional frailty.4,58 Because only one study using sarcopenia as a marker of frailty was available for our primary analyses, a sensitivity analysis would likely not add value, but future evaluation may be required in prospective studies. Finally, because we relied on observational studies, our findings reflect associations that we cannot prove to be fully causal. There may yet be underlying confounders that this review has not been able to capture, such as disproportionate anticoagulant medication use.59


In a systematic review and meta-analysis of frailty and outcomes after multisystem trauma, we found evidence of an association between frailty and mortality, as well as strong associations between frailty and complications and adverse discharge disposition; LoS was also increased for people with frailty. Future research is needed to better understand the association between frailty and other important patient- and caregiver-centered outcomes, as well as to continue to develop interventions to help improve outcomes for older trauma patients with frailty.


All authors acknowledge the support of Sasha Davis (the Ottawa Hospital Information Services) for her assistance in developing the search strategy, and the Ottawa Hospital Department of Anesthesiology for support of Distiller SR licenses.


Name: Alexander Poulton, MD.

Contribution: This author helped with conception, design, data acquisition, analysis, and interpretation, and drafted, revised, and approved the final manuscript.

Name: Julia F. Shaw, BSc.

Contribution: This author helped with design, data acquisition, analysis, and interpretation, and drafted, revised, and approved the final manuscript.

Name: Frederic Nguyen, MD.

Contribution: This author helped with design, data acquisition, analysis, and interpretation, and drafted, revised, and approved the final manuscript.

Name: Camilla Wong, MD, FRCPC.

Contribution: This author helped with design, analysis, and interpretation, and drafted, revised, and approved the final manuscript.

Name: Jacinthe Lampron, MD, FRCSC.

Contribution: This author helped with design, analysis, and interpretation, and drafted, revised, and approved the final manuscript.

Name: Alexandre Tran, MD, MSc.

Contribution: This author helped with design, analysis, and interpretation, and drafted, revised, and approved the final manuscript.

Name: Manoj M. Lalu, MD, PhD, FRCPC.

Contribution: This author helped with design, analysis, and interpretation, and drafted, revised, and approved the final manuscript.

Name: Daniel I. McIsaac, MD, MPH, FRCPC.

Contribution: This author helped with conception, design, data acquisition, analysis, and interpretation, and drafted, revised, and approved the final manuscript and is guarantor.

This manuscript was handled by: Robert Whittington, MD.



    1. Vincent GK, Velkoff VA. The next four decades: the older population in the United States: 2010 to 2050. 2010. Washington, DC: US Census Bureau; Available at: Accessed January 24, 2020.
    2. Chehade M, Gill TK, Visvanathan R. Low energy trauma in older persons: where to next? Open Orthop J. 2015;9:361–366.
    3. Chang MC. National Trauma Data Bank (NTDB): Annual Report 2016. Available at: Accessed January 24, 2020.
    4. Fried LP, Tangen CM, Walston J, et al.; Cardiovascular Health Study Collaborative Research Group. Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci. 2001;56:M146–M156.
    5. Rockwood K, Song X, MacKnight C, et al. A global clinical measure of fitness and frailty in elderly people. CMAJ. 2005;173:489–495.
    6. Turner G, Clegg A; British Geriatrics Society; Age UK; Royal College of General Practioners. Best practice guidelines for the management of frailty: a British Geriatrics Society, Age UK and Royal College of General Practitioners report. Age Ageing. 2014;43:744–747.
    7. Lin HS, Watts JN, Peel NM, Hubbard RE. Frailty and post-operative outcomes in older surgical patients: a systematic review. BMC Geriatr. 2016;16:157.
    8. Theou O, Squires E, Mallery K, et al. What do we know about frailty in the acute care setting? A scoping review. BMC Geriatr. 2018;18:139.
    9. Theou O, Brothers TD, Mitnitski A, Rockwood K. Operationalization of frailty using eight commonly used scales and comparison of their ability to predict all-cause mortality. J Am Geriatr Soc. 2013;61:1537–1551.
    10. McDonald VS, Thompson KA, Lewis PR, Sise CB, Sise MJ, Shackford SR. Frailty in trauma: a systematic review of the surgical literature for clinical assessment tools. J Trauma Acute Care Surg. 2016;80:824–834.
    11. Cubitt M, Downie E, Shakerian R, Lange PW, Cole E. Timing and methods of frailty assessments in geriatric trauma patients: a systematic review. Injury. 2019;50:1795–1808.
    12. Llompart-Pou JA, Pérez-Bárcena J, Chico-Fernández M, Sánchez-Casado M, Raurich JM. Severe trauma in the geriatric population. World J Crit Care Med. 2017;6:99–106.
    13. Publisher Institute for Healthcare Improvement. Triple Aim for Populations. Available at: Accessed January 24, 2020.
    14. Berwick DM, Nolan TW, Whittington J. The triple aim: care, health, and cost. Health Aff (Millwood). 2008;27:759–769.
    15. Stiefel M, Nolan K, Steifel M NK. A guide to measuring the triple aim: population health, experience of care, and per capita cost. 2012.Cambridge, MA: Institute for Healthcare Improvement.
    16. Stroup DF, Berlin JA, Morton SC, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA. 2000;283:2008–2012.
    17. Higgins JGS, Higgins JP, Green S. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0, 5.1. 2011.London, UK: Cochrane Collaboration.
    18. Riley RD, Moons KGM, Snell KIE, et al. A guide to systematic review and meta-analysis of prognostic factor studies. BMJ. 2019;364:k4597.
    19. Moher D, Liberati A, Tetzlaff J, Altman DG; PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6:e1000097.
    20. McGowan J, Sampson M, Salzwedel DM, Cogo E, Foerster V, Lefebvre C. PRESS peer review of electronic search strategies: 2015 guideline statement. J Clin Epidemiol. 2016;75:40–46.
    21. Maxwell CA, Mion LC, Mukherjee K, et al. Preinjury physical frailty and cognitive impairment among geriatric trauma patients determine postinjury functional recovery and survival. J Trauma Acute Care Surg. 2016;80:195–203.
    22. Akpan A, Roberts C, Bandeen-Roche K, et al. Standard set of health outcome measures for older persons. BMC Geriatr. 2018;18:36.
    23. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7:177–188.
    24. Wan X, Wang W, Liu J, Tong T. Estimating the sample mean and standard deviation from the sample size, median, range and/or interquartile range. BMC Med Res Methodol. 2014;14:135.
    25. Sterne JA, Sutton AJ, Ioannidis JP, et al. Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials. BMJ. 2011;343:d4002.
    26. Egger M, Smith GD, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315:629–634.
    27. Sterne JA, Hernán MA, Reeves BC, et al. ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions. BMJ. 2016;355:i4919.
    28. Kaplan SJ, Pham TN, Arbabi S, et al. Association of radiologic indicators of frailty with 1-year mortality in older trauma patients: opportunistic screening for sarcopenia and osteopenia. JAMA Surg. 2017;152:e164604.
    29. Joseph B, Pandit V, Rhee P, et al. Predicting hospital discharge disposition in geriatric trauma patients: is frailty the answer? J Trauma Acute Care Surg. 2014;76:196–200.
    30. Min L, Ubhayakar N, Saliba D, et al. The vulnerable elders survey-13 predicts hospital complications and mortality in older adults with traumatic injury: a pilot study. J Am Geriatr Soc. 2011;59:1471–1476.
    31. Cheung A, Haas B, Ringer TJ, McFarlan A, Wong CL. Canadian study of health and aging clinical frailty scale: does it predict adverse outcomes among geriatric trauma patients? J Am Coll Surg. 2017;225:658.e3–665.e3.
    32. Joseph B, Orouji Jokar T, Hassan A, et al. Redefining the association between old age and poor outcomes after trauma: the impact of frailty syndrome. J Trauma Acute Care Surg. 2017;82:575–581.
    33. Joseph B, Phelan H, Hassan A, et al. The impact of frailty on failure-to-rescue in geriatric trauma patients: a prospective study. J Trauma Acute Care Surg. 2016;81:1150–1155.
    34. Joseph B, Pandit V, Zangbar B, et al. Validating trauma-specific frailty index for geriatric trauma patients: a prospective analysis. J Am Coll Surg. 2014;219:10.e1–17.e1.
    35. Ebbeling L, Grabo DJ, Shashaty M, et al. Psoas:lumbar vertebra index: central sarcopenia independently predicts morbidity in elderly trauma patients. Eur J Trauma Emerg Surg. 2014;40:57–65.
    36. Curtis E, Romanowski K, Sen S, Hill A, Cocanour C. Frailty score on admission predicts mortality and discharge disposition in elderly trauma patients over the age of 65 y. J Surg Res. 2018;230:13–19.
    37. Hamidi M, Zeeshan M, O’Keeffe T, et al. Prospective evaluation of frailty and functional independence in older adult trauma patients. Am J Surg. 2018;216:1070–1075.
    38. Lauerman MH, Raithel M, Kufera J, et al. Comparison of individual and composite radiographic markers of frailty in trauma. 2019;50:149–155.
    39. Mccusker A, Khan M, Kulvatunyou N, et al. Sarcopenia defined by a computed tomography estimate of the psoas muscle area does not predict frailty in geriatric trauma patients. Am J Surg. 2019;218:261–265.
    40. Tipping CJ, Hodgson CL, Harrold M, Chan T, Holland AE. Frailty in patients with trauma who are critically ill: a prospective observational study to determine feasibility, concordance, and construct and predictive validity of 2 frailty measures. Phys Ther. 2019;99:1089–1097.
    41. Couch AP, Ho K, Darwood R, Hsu JM. Lean psoas area does not correlate with clinical outcomes in moderately to severely injured older people. Australas J Ageing. 2018;37:E7–E11.
    42. Joseph B, Pandit V, Zangbar B, et al. Superiority of frailty over age in predicting outcomes among geriatric trauma patients: a prospective analysis. JAMA Surg. 2014;149:766–772.
    43. Lenartowicz M, Parkovnick M, McFarlan A, et al. An evaluation of a proactive geriatric trauma consultation service. Ann Surg. 2012;256:1098–1101.
    44. Kim DH, Kim CA, Placide S, Lipsitz LA, Marcantonio ER. Preoperative frailty assessment and outcomes at 6 months or later in older adults undergoing cardiac surgical procedures: a systematic review. Ann Intern Med. 2016;165:650–660.
    45. Darvall JN, Gregorevic KJ, Story DA, Hubbard RE, Lim WK. Frailty indexes in perioperative and critical care: a systematic review. Arch Gerontol Geriatr. 2018;79:88–96.
    46. Watt J, Tricco AC, Talbot-Hamon C, et al. Identifying older adults at risk of harm following elective surgery: a systematic review and meta-analysis. BMC Med. 2018;16:2.
    47. Bagshaw SM, Stelfox HT, McDermid RC, et al. Association between frailty and short- and long-term outcomes among critically ill patients: a multicentre prospective cohort study. CMAJ. 2014;186:E95–E102.
    48. Mangram AJ, Mitchell CD, Shifflette VK, et al. Geriatric trauma service: a one-year experience. J Trauma Acute Care Surg. 2012;72:119–122.
    49. Olufajo OA, Tulebaev S, Javedan H, et al. Integrating geriatric consults into routine care of older trauma patients: one-year experience of a level i trauma center. J Am Coll Surg. 2016;222:1029–1035.
    50. Sharma J, Parulekar M, Stewart P, et al. Geriatric consultation reduces high-risk medication usage at discharge in elderly trauma patients. Cureus. 2018;10:e3649.
    51. Harari D, Hopper A, Dhesi J, Babic-Illman G, Lockwood L, Martin F. Proactive care of older people undergoing surgery (‘POPS’): designing, embedding, evaluating and funding a comprehensive geriatric assessment service for older elective surgical patients. Age Ageing. 2007;36:190–196.
    52. McDonald SR, Heflin MT, Whitson HE, et al. Association of integrated care coordination with postsurgical outcomes in high-risk older adults: the Perioperative Optimization of Senior Health (POSH) Initiative. JAMA Surg. 2018;153:454–462.
    53. Moore L, Stelfox HT, Turgeon AF, et al. Rates, patterns, and determinants of unplanned readmission after traumatic injury: a multicenter cohort study. Ann Surg. 2014;259:374–380.
    54. Fried LP, Ferrucci L, Darer J, Williamson JD, Anderson G. Untangling the concepts of disability, frailty, and comorbidity: implications for improved targeting and care. J Gerontol A Biol Sci Med Sci. 2004;59:255–263.
    55. McIsaac DI, Taljaard M, Bryson GL, et al. Frailty as a predictor of death or new disability after surgery: a Prospective Cohort Study. Ann Surg. 2020;271:283–289.
    56. Shears M, Takaoka A, Rochwerg B, et al.; Canadian Critical Care Trials Group. Assessing frailty in the intensive care unit: a reliability and validity study. J Crit Care. 2018;45:197–203.
    57. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003;327:557–560.
    58. Cesari M, Landi F, Vellas B, Bernabei R, Marzetti E. Sarcopenia and physical frailty: two sides of the same coin. Front Aging Neurosci. 2014;6:192.
    59. Dossett LA, Riesel JN, Griffin MR, Cotton BA. Prevalence and implications of preinjury warfarin use: an analysis of the National Trauma Databank. Arch Surg. 2011;146:565–570.

    Supplemental Digital Content

    Copyright © 2020 International Anesthesia Research Society