What is previously known

  • Excess mortality persists for years post hip fracture.

  • Several patient and system factors have been associated with the risk of death in patients with hip fracture.

  • However underlying mechanisms of the found associations are rarely discussed.

What this study adds

  • We synthesized the evidence from 56 recent articles that reported on 35 patient and 9 system factors of mortality post hip fracture.

  • The majority of factors were reported with no proposed mechanisms for their effects on mortality. Where reported, underlying mechanisms are often based on a single article.

  • The proposed biological mechanisms include complications, comorbidity, cardiorespiratory function, immune function, bone remodeling and glycemic control.

Background

Hip fracture is a leading cause of injury related mortality in older adults [1]. Omsland and colleagues reported mortality rates five times higher in men and three times higher in women compared to the general population in the first year post fracture [2]. This excess mortality persists 10 years post fracture [2, 3].

The existing literature identifies patient and system factors associated with the risk of death among patients with hip fracture. However, no attempt has been made to synthesize this literature on the underlying mechanisms for these associations.

Without knowledge of mechanisms mediating an association, interventions to improve survival post hip fracture can only be designed on the basis of the reported statistical associations. We believe designing interventions should rely on knowledge about a modifiable factor with negative effect on survival. Where modifiable factors lie on the causal pathway between exposure and outcome determines the focus of an intervention.

Scoping reviews represent an approach to summarizing the range of evidence on a subject, to clarify a complex concept, and to help refine subsequent research questions for a full systematic review [4, 5]. This review contributes to the existing literature by synthesizing the evidence available on patient and system factors of mortality after hip fracture. To go beyond traditional reports, we extract and synthesize additional information on biological and hypothetical mechanisms for reported associations. More specifically, the aims of this review are 1) to identify patient and system factors of mortality after hip fracture, and 2) collate the description of proposed mechanisms for their associations with mortality.

Methods

The key elements of the scoping review framework include formulating the research question, identifying relevant studies, selecting studies from electronic database, charting the extracted data and collating, summarizing and reporting findings. We extend this framework by collecting information on the underlying mechanisms for found associations. We have followed to a widely recognized framework by Arksey and O’Malley [6] and recommendations of Levac et al. [5] for conducting and reporting scoping reviews. This scoping review synthesizes published literature and ethical approval was not required.

The population of interest is frail adults aged 50 years or older admitted to acute care with non-pathologic low energy hip fracture. Concepts of interest include both patient and system factors. The outcome of interest is mortality following usual care.

Study selection

One reviewer searched MEDLINE using the search terms “mortality” Medical Subject Headings [MeSH] AND “hip fracture” [MeSH] and screened studies for eligibility. Studies were first screened according to title and abstract with those that appeared suitable selected for a full-text review using standardized inclusion criteria (Table 1). Studies marked as ‘maybe for inclusion’ were screened by a second reviewer for eligibility.

Table 1 Inclusion criteria for the literature search
Full size table

We included reports from 2009 or later to minimize the potential biasing effects of demographic aging [7–9], surgical advancements [10], and changes in delivery of hip fracture care [11–13]. Intervention-based studies were excluded as they do not reflect hip fracture mortality following usual care. Studies whose main independent variables were laboratory tests or operation type were also excluded as they were considered outside of the current scope of interest. Finally, only studies which conducted a regression analysis were included as regression analysis was deemed a proxy for adequate sample size [14].

Using a formal instrument, one reviewer extracted authors name, publication date, timing of assessment relative to the hip fracture event, length of follow up, patient and system factors from each article. The significance of statistical associations between the factors and mortality was derived from the 95 % confidence intervals reported in the articles. The proposed mechanisms for mortality were extracted from discussions by one reviewer. The accuracy of extraction was assessed by a second reviewer.

Collating, summarizing and reporting results

Patient and system factors of mortality studied in the reviewed articles are summarized in Tables 2, 3 and 4. Factors with a proposed mechanism of their effects on mortality are summarized in Tables 5 and 6 with indication whether the mechanisms is hypothetical.

Table 2 Articles studying mortality in relation to injury and complications
Full size table
Table 3 Articles studying mortality in relation to demographic factors and comorbidity
Full size table
Table 4 Articles studying mortality in relation to system factors
Full size table
Table 5 Proposed mechanisms and mediators for the mortality effect of patient factors
Full size table
Table 6 Proposed mechanisms and mediators for the effect of system factors on mortality
Full size table

Results

Search results

The search produced 241 articles for initial title and abstract screening. Figure 1 shows the selection process which identified 56 articles used in this review. Among the selected articles, 21 reported on in-hospital mortality [15–35], 4 reported on 30 day mortality [36–39], 20 reported on 12 months mortality [40–59], and 11 reported more than 12 month mortality [60–70].

Fig. 1
figure 1

Flow chart of the literature retrieval, review, exclusion and selection with sorting by follow up time. n = number. * = Studies excluded with patient populations less than 50 years old, pathological or high impact hip fractures, or whose main independent variables were laboratory tests or operation type

Full size image

Patient factors of mortality

We identified 35 patient factors of mortality post hip fracture reported in the reviewed articles, Tables 2 and 3. The majority of factors were studied by only one or two studies included in this review. There is a general consensus in the literature that mortality is associated with age, sex, comorbidity, functional status, dementia, arrhythmia and congestive heart failure. We noted conflicting reports for the association between mortality and both fracture type [17, 65] and delirium [47, 62, 66].

For 14 factors we found a proposed mechanism of their effects on mortality, Table 5. Biological mechanisms included comorbidity [47, 70], cardiorespiratory function [68], immune function [38], bone remodeling [52], glycemic control [52], and calcium homeostasis [32]. Non-biological mechanisms included hospitalization delay [50], surgical delay [25, 48] and length of stay [36]. Some proposed included hypothetical mediators, such as reduced reserve capacity [22, 37], a patent foramen ovale [37] and reduced wound healing [52]. Fig. 2 shows two examples of the mechanisms proposed for patient factors in the reviewed articles. First, the onset of complications mediates the effect of cardiorespiratory function on mortality [63]. Second, a hypothetical reduction in reserve capacity mediates the mortality effect of age and extent of comorbidity [22, 37].

Fig. 2
figure 2

Examples of mechanisms proposed for patient factors in reviewed articles. Black node indicates the outcome. White nodes indicates a reported factor. Square box indicates a measurable mediator. Grey node indicates a hypothetical mediator

Full size image

System factors of mortality

In the reviewed articles, we identified 9 system factors of mortality post hip fracture including hospitalization delay, July admission, surgical delay, anaesthetic type, intensive care admission, hospital volume, surgeon volume, nursing volume and length of stay, Table 4. There is no consensus in the literature on system factors of mortality. The most studied factor was surgical delay (9 articles). However, the association of mortality with surgical delays is disputed by reports of no association [15, 47, 49, 50]. We also noted conflicting reports for the association between mortality and both July admission and hospital volume. The other factors were studied by only one or two studies included in this review.

For 7 factors we found a proposed mechanism of their effects on mortality, Table 6. Complications were proposed as a biological mechanism for the mortality effect of nursing staff volume [28] and hospitalization delay [50, 61]. Non-biological mechanisms included surgical delay [28], staffing volume [31] and discharge destination [36]. Some proposed included hypothetical mediators, such as, exposure to inflammatory and hypercoagulable states [71, 72] and inappropriate planning, technique or management [34].

Discussion

The purpose of this review was to synthesize the information available on proposed mechanisms for reported associations between patient and system factors and mortality after hip fracture. The articles included in this review point to plausible mediators in the biological mechanisms for mortality post fracture: complications, comorbidity, cardiorespiratory function, immune function, bone remodeling and glycemic control. For example, exposure to immobilization and inflammatory states is the proposed mechanism mediating the mortality effect of hospitalization delay [71, 72]. As argued elsewhere, prolonged immobilization leads to potentially fatal complications such as pulmonary embolism and pneumonia while prolonged hypercoagulable inflammation leads to potentially fatal complications including stroke and myocardial infarction [73].

A hypothetical reduction in reserve capacity, whereby a patient cannot withstand the stress of trauma and their pre-existing comorbidity [38], was proposed as a mechanism for the mortality effect of comorbidity [22, 37]. It seems plausible, because numerous studies associated mortality with coexisting arrhythmia, congestive heart failure, coronary artery disease, myocardial infarction, anemia and cerebrovascular accident. As noted elsewhere, patients undergoing hip fracture surgery require the reserve capacity to withstand the cardiovascular depressant effect of anaesthesia [74]. For those who survive beyond the short-term, patients with cardiovascular disease more often present with reduced reserve exercise capacity [75] compromising their rehabilitation potential and placing them at greater risk of dependency, complications and death [76].

This is the first scoping review to synthesize the proposed biological and hypothetical mechanisms for patient and system factors of mortality following hip fracture. Such synthesis represents a first step towards transparency about underlying assumptions when informing policy on potential interventions to improve survival in this vulnerable population.

This review is not without limitations. In contrast to a systematic review, where literature is critically appraised on the methodology, we assess the reviewed articles only according to the presence of proposed mechanisms for the reported associations. This is a common approach in scoping reviews where the purpose is to collate the evidence on a topic of interest. [5] The search strategy was restricted to one database over a 5 year period preceding the review development to minimize the potential biasing effects of surgical advancements [10], and changes in delivery of hip fracture care [11–13]. We excluded articles reporting outcomes of interventions as they do not reflect hip fracture mortality resulting from usual care. These restrictions may result in lacking some articles both on factors of mortality and proposed mechanisms.

Conclusions

We synthesized proposed mechanisms for reported associations between patient and system factors and mortality after hip fracture. We identified complications, comorbidity, cardiorespiratory function, immune function and bone remodeling and glycemic control as plausible mediators in the biological mechanisms for mortality post fracture. However, we found that the majority of patient and system factors of mortality post hip fracture were reported by only one or two articles and with no proposed mechanisms for their effects on mortality. Where reported, underlying mechanisms are often based on a single article and should be confirmed with further study. Therefore, one cannot be certain whether intervening on such factors may produce expected results.

Ethical approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Availability of data and materials

Articles were identified in MEDLINE. All articles included in the final review are identified in the reference list. Access to full text for each article is dependent on journal and institutional constraints.