Mortality After Hip and Spine Fractures in Patients With End-Stage Kidney Disease: A Systematic Review and Meta-Analysis

Fractures represent a major cause of disability in the elderly, and patients with fractures exhibit a higher mortality rate than those without. Fractures are also an important health problem among patients with end-stage kidney disease (ESKD) requiring hemodialysis, peritoneal dialysis, or kidney transplantation. To the best of our knowledge, no study in the literature has yet quantitatively summarized the mortality rates, and a summary of evidence on post-hip and spine fracture mortality in patients with ESKD is lacking. The purpose of this study is to quantitatively evaluate the mortality rate, one-year mortality rate, and five-year mortality rate after hip and spine fractures in patients with ESKD receiving kidney replacement therapy. The MEDLINE, Cochrane Central Register of Controlled Trials (CENTRAL), Embase, and ClinicalTrials.gov databases were comprehensively searched for reports on mortality rate and time-period mortality in patients with ESKD after hip or spine fractures up to June 2022. Prospective and retrospective cohort studies, as well as case series involving four or more patients, were included. Pooled mortality rate, one-year rate, and five-year mortality rate with 95% confidence intervals (CIs) were examined using a random-effects model. The risk of bias was assessed using the Joanna Briggs Institute (JBI) Prevalence Critical Appraisal Tool. Additionally, heterogeneity between studies was evaluated. A total of 26 studies were included in this meta-analysis. The one-year and five-year mortality rates after hip and spine fractures were 215.35-774.0 per 1,000 person-year and 148-194.1 per 1,000 person-year, respectively. After hip fractures, the one-year mortality rate was 27% (95% CI: 18-38%, I2 = 98%), whereas the five-year mortality rate was 56% (95% CI: 41-71%, I2 = 99%). After spine fractures, the one-year mortality rate was 10% (95% CI: 4-17%, I2 = 70%), whereas the five-year mortality rate was 48.3%. The post-fracture mortality rate was high in patients with ESKD, particularly within one year after the occurrence of fractures. Additionally, the five-year mortality rate after hip femoral or spine fractures was high at approximately 50%.


Introduction And Background
Fractures represent a major cause of disability in the elderly, and the risk of fractures, including inadvertent falls, frailty, osteoporosis, and menopause, increases with age [1].More than two million osteoporosisrelated fractures have been estimated to occur in the United States, with spine and hip fractures accounting for 27% and 14%, respectively [2].Patients with fractures exhibit a higher mortality rate than those without [3][4][5].In particular, the one-year mortality rate after hip fractures has been reported to be approximately 3.7 and 2.8 times higher in men and women with fractures, respectively, than in non-fracture patients [5].
Fractures are also an important health problem among patients with end-stage kidney disease (ESKD) requiring hemodialysis, peritoneal dialysis, or kidney transplantation [6,7].Compared to the general population, patients with ESKD have a two-to four-fold increased risk of hip and spine fractures [8][9][10][11].Such an excessive fracture risk in patients with ESKD is likely attributable to underlying mineral metabolism abnormalities that lead to renal osteodystrophy, as well as an increased fall risk from neuromuscular impairments [12].In particular, hip fracture has severe consequences in patients with ESKD, including an increased risk of hospitalization, reduced quality of life, loss of independence, and death [12][13][14][15][16][17].
Previous cohort studies have reported one-year mortality rates ranging from 14% to 43% after hip fractures in patients with ESKD [15][16][17]; however, the reported rates widely vary across studies, possibly owing to differences in sample sizes, patient demographics, or kidney replacement therapy modalities.To the best of our knowledge, no study in the literature has yet quantitatively summarized the mortality rates, and a summary of evidence on post-spine fracture mortality in patients with ESKD is lacking.Therefore, we conducted a systematic review and meta-analysis to provide precise estimates of mortality after hip and spine fractures in patients with ESKD who are undergoing kidney replacement therapy.

Compliance With Reporting Guidelines
This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [18].We confirmed that our systematic review was PRISMA-compliant by consulting the 2020 PRISMA checklist (Appendix Table 3).The prespecified protocol can be accessed at https://www.protocols.io/view/systematic-review-and-meta-analysis-of-incidence-a-capzsdp6.While this study involved human participants, ethical approval from the institutional review board was not obtained, given that ethical approval was sought by the individual original studies included in the systematic review.Informed consent was obtained from all individual participants included in each study prior to their study participation.All authors received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors for this research.

Eligibility Criteria for the Included Studies
The eligibility criteria were as follows: (ⅰ) Study design: prospective and retrospective cohort studies and case series involving four cases or more (a case referred to a patient with ESKD after hip or spine fracture) (ⅱ) Study population: patients with ESKD, defined as the requirement for hemodialysis, peritoneal dialysis, or kidney transplantation, irrespective of primary disease (ⅲ) Outcome: studies reporting on mortality or mortality rate (ⅳ) Time: outcome reported at least one month after hip or spine fracture Studies were eligible irrespective of publication status, follow-up period, language, age, sex, race, or surgery status.However, studies that did not recruit or were withdrawn from ClinicalTrials.gov,case reports describing three or fewer cases, animal and laboratory studies, and literature reviews were excluded.

Outcome Measures
The primary outcomes evaluated were the (i) mortality rates after hip and spine fractures, (ii) one-year mortality rate after hip and spine fractures, and (iii) five-year mortality rates after hip and spine fractures.The diagnoses of hip and spine fractures set by the original authors included ascertainment from the presence of a corresponding International Classification of Diseases, 9th Revision, code in a hospital billing claim.

Search Methods for Study Identification
Electronic searches: The Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE (via Ovid), and Embase (via ProQuest) databases were searched for relevant studies on June 23, 2022.The search results were filtered for the prognostic factors reported by Wilczynski et al. [19] (Appendix Tables 4-8).
Searches of other resources: The World Health Organization's International Clinical Trials Registry Platform (ICTRP) Search Portal and ClinicalTrials.govregistry were also searched to identify completed unpublished studies and to investigate reporting bias.Furthermore, the references of extracted studies and international guidelines were checked, and the authors were contacted if the extracted studies lacked the necessary data.The detailed search strategies are described in Appendix Tables 7, 8.

Data Collection and Analysis
Study selection: Two out of the four reviewers, Yoshinosuke Shimamura (Y.S.), Hiroshi Ueta (H.U.), Takamasa Miyauchi (T.M.), and Mari Yamamoto (M.Y.), independently screened the titles and abstracts identified during the search.A predefined protocol was followed in screening the abstracts and full texts, and predefined criteria were used in the registered protocol.All extracts from the reviewers were subjected to a full-text review; subsequently, they independently determined whether the full text should be included in the review.The first author (Y.S.) checked all included studies and applied the exclusion criteria for all records subjected to the full-text screening procedure; hence, the decision did not differ systematically.The original authors were contacted if the study had an abstract only or if it was unclear whether the study met the review criteria.Any disagreement was resolved through discussion between the two reviewers; if an agreement could not be reached, a third reviewer, Yasutaka Kuniyoshi (Y.K.) or Yasushi Tsujimoto (Y.T.), acted as an arbiter.
Data extraction and management: Two reviewers independently performed the data extraction, and any disagreement between the two reviewers was resolved through discussion.A third reviewer was involved in the discussion, where necessary, and the original authors were contacted.The one-year and five-year mortality rates were extracted; conversely, mortality rates at less than one year were not extracted.Additionally, data were extracted when the studies reported mortality rates at more than one year and less than five years (e.g., two-year mortality rate), even if they did not report either a one-year or five-year mortality rate.In such cases, data on the outcome at less than three years were considered as the one-year mortality rate, whereas data on the outcome at three to five years were considered as the five-year mortality rate.A pre-checked data extraction form with 10 randomly selected studies was utilized.The mortality rate was calculated as described in respective studies.
Assessment of the risk of bias of included studies: Two reviewers independently assessed the risk of bias in each study using the JBI Prevalence Critical Appraisal Tool [20,21].The following domains were assessed: 1. Was the sample frame appropriate to address the target population?2. Were the study participants sampled appropriately?
3. Was the sample size adequate?4. Were the study participants and setting described in detail? 5. Was the data analysis conducted with sufficient coverage of the identified sample?6. Were valid methods used for the identification of the condition?7. Was the condition measured in a standard and reliable manner for all participants?8. Was there appropriate statistical analysis?9. Was the response rate adequate?If not, was a low response rate managed appropriately?Any disagreement was resolved through discussion among the reviewers; if an agreement could not be reached, a third reviewer acted as an arbiter.In this study, the overall risk of bias was calculated as the number of "yes" responses for each domain divided by the total number of domains and was expressed as a percentage.The overall risk of bias was interpreted according to the calculated percentage as follows: <50%, high risk of bias; 50-80%, moderate risk of bias; >80%, low risk of bias [21].

Measures of the Treatment Effect
In this study, both the incidence rate (measured as the number of incident cases per measure of exposure) and incidence proportion (measured as the number of incident cases over a specified period) were determined with 95% confidence intervals (CIs) and 95% prediction intervals.The between-study variance was estimated using tau2 statistics, which supply a logit scale measure of between-study variance, represented in a more readily interpretable way by 95% prediction intervals.

Data Synthesis
A single-arm analysis was conducted.Percentages, means, and standard deviations were calculated for categorical variables.The pooled mortality rate and mortality were calculated for patients with ESKD after hip fractures and those with ESKD after spine fractures.A random-effects model (DerSimonian and Laird approach) was used for pooled estimates to consider the variance between and among the studies.Statistical analyses were performed using R software (R Development Core Team 2019), with meta version 4.15-0 and metaphor version 2.4-0.For dropouts, imputation was not performed in accordance with the recommendations by The Cochrane Handbook [22].A meta-analysis was conducted on data presented by the original authors, and any missing values or summary statistics were not complemented.

Subgroup Analysis
Considering that the present study aimed to identify the plausible causes of heterogeneity, the following prespecified subgroup analyses of primary outcomes were planned: sex (men vs. women), race (Black vs. non-Black), presence or absence of cardiovascular diseases (hypertension, coronary artery disease, congestive heart failure, cerebrovascular disease, and peripheral vascular disease), presence or absence of diabetes mellitus, type of kidney replacement therapy (hemodialysis, peritoneal dialysis, and kidney transplantation), and participants' age category (≥75 years vs. <75 years).

Sensitivity Analysis
To confirm the robustness of the main results, a prespecified sensitivity analysis was conducted on the primary outcomes, excluding the outcome data other than the one-year and five-year mortality rates.

Assessment of Reporting Bias
The ICTRP and ClinicalTrials.govwere searched for studies that were completed but have not yet been published.Potential publication bias was assessed via visual inspection of funnel plots and Egger's test.

Literature Search
After duplicate removal, a total of 657 records were identified through a systematic search in MEDLINE, Embase, CENTRAL, ClinicalTrials.gov,and ICTRP.Among these, 69 reports were retrieved for full-text review; however, five reports were excluded because of duplicate publications, resulting in 64 eligible reports based on the inclusion criteria.Furthermore, seven studies with incorrect study designs, 19 studies with incorrect study populations, three studies with incorrect interventions, and seven studies with incorrect outcomes were excluded.Notably, the study by Yuan et al. [23] was excluded from the main analysis because it was a letter article, and the single-center cohort study by Iseri et al. [24] was also excluded because it primarily involved patients with malnutrition, inflammation, and atherosclerosis syndrome and did not provide details on the status of hip and spine fractures in participants.Finally, 26 studies were included in the meta-analysis (Figure 1).

Discussion
By performing a systematic review and meta-analysis, we comprehensively and quantitatively analyzed the mortality rates and mortality after hip and spine fractures in patients with ESKD receiving kidney replacement therapy.We found that when patients with ESKD had hip and spine fractures, their mortality rate was high, particularly during the first year after these fractures.Notably, 27% and 56% of patients with ESKD after hip fractures died at one year and five years later, respectively.These results suggest that patients and their family members should be informed of poor prognosis, and clinicians should strictly and appropriately manage the mineral bone disease.Such a high risk of post-fracture mortality among patients with ESKD is consistent with the findings of a previous study by Tentori et al. [13], who utilized data from the Dialysis Outcomes and Practice Patterns Study and reported that the mortality rate was the highest in the first month following the fracture event and that the rate declined thereafter.
There are several explanations for the higher risk of post-fracture mortality in patients with ESKD than those in the general population.First, patients with ESKD might have been in such a poor condition at the time of fracture occurrence that they were judged intolerable to surgery.Indeed, when patients with ESKD sustain hip fractures, conservative management is chosen in up to 13% of patients [14].In the general population, however, most patients are treated with surgical repair or arthroplasty [47,48], and nonoperative management is selected for only 2.6% of patients after hip fractures.Furthermore, a previous study [15] using data from the United States Renal Data System showed that patients with fractures who did not undergo surgery at the time of their fracture had a higher all-cause mortality than those who did.This applies to patients without ESKD.A previous study [49] showed that geriatric hip fracture patients who were treated nonoperatively had a higher in-hospital, 30-day, and one-year mortality than a matched group of operatively treated patients.Hence, nonoperative management of hip fractures in patients with ESKD may be related to the decline in mobility and muscle strength, contributing to the high mortality rate [49,50].Another possible explanation is that patients with ESKD after fractures may experience postoperative complications.The results of several previous studies support this hypothesis by showing that patients undergoing dialysis have an increased risk of infectious and cardiovascular complications after surgery [51,52].For example, Benjamin et al. [51] reported that patients on dialysis had 2.9-and 1.7-fold higher risks of sepsis and pneumonia, respectively, than non-dialyzed patients within 30 days of hip fracture fixation.Additionally, a previous meta-analysis [52] highlighted that patients on dialysis had a two-to fivefold increase in the odds of postoperative myocardial infarction and stroke, regardless of the surgical procedure.Alternatively, this result may be confounded by indications, and patients who do not undergo surgery are less likely to undergo surgery because of poor health status or comorbidities at the time of fracture occurrence.
We also found that the one-year and five-year mortality rates were 10% and 48%, respectively, when patients with ESKD sustained spine fractures.To our knowledge, this is the first study to systematically evaluate the prognosis after spine fractures in patients with ESKD; however, our findings should be interpreted with caution because the five-year mortality rate solely depends on the results of a study by Maeno et al. [42].Our results showed that many previous studies on the prognosis of post-spine fractures were limited to short-term observations, and they have reported that hyperkyphosis and vascular calcifications were associated with poor prognosis in patients after spine fracture [30,37,53].For instance, a prospective multicenter cohort study [30] following incident dialysis for one year reported that thoracic hyperkyphosis and increased curvature of the thoracic spine were associated with a higher risk of all-cause mortality.Moreover, another single-center cohort study [53] showed an association between vertebral fractures (thoracic and lumbar) and higher two-year mortality in female patients receiving hemodialysis.Furthermore, the results of these studies [30,37,53] did not report the cause of death; therefore, it is unclear whether the deaths were due to cardiovascular diseases or other causes.The longitudinal influence of spine fractures on mortality should be further investigated with a focus on the cause of death in future studies.
However, this study had several limitations.First, there may have been some missing studies because we included the outcome term in our search.Because database searches can only identify search terms in titles and abstracts, the inclusion of outcome terms may be subject to outcome reporting bias [54].Second, we only included studies written in English.Nonetheless, this study included studies reported from Europe, the United States, Canada, and Asian countries, which we believe minimizes selection bias and ensures generalizability.Third, differences in the study design, participant populations, and diagnostic procedure likely contributed to the heterogeneity of our results.Indeed, we could not perform several preplanned subgroup analyses to explain the heterogeneity of our findings, and only a few studies were included.Also, it should be noted that diagnostic procedures for spine fractures vary in the respective articles [30,37,53].Additionally, several of the included studies did not report important clinical information, such as dialysis vintage, history of cardiovascular diseases, diabetes, or past fractures.This may affect the reliability with which clinicians can apply our estimates.Fourth, several studies involved the same investigators, and it was not possible to completely exclude overlaps in the patient population.Fifth, hip and spine fractures may not be comparable because of differences in treatment and prognosis, but these two conditions are common in patients with ESKD.

Conclusions
In conclusion, this study showed that patients with ESKD sustaining hip and spine fractures had high postfracture mortality rates.The one-year mortality was 27% after hip fracture in 14 studies and 10% after spine fracture in two studies.Additionally, the five-year mortality rate was 56% after hip fractures in six studies and 48% after spine fractures in one study.While it remains unclear whether deaths were related to fractures or a consequence of cardiovascular diseases or infections, understanding the prognosis of these types of fractures will contribute to guiding clinical management and fracture prevention in this high-risk population.More importantly, it is essential in advancing care planning and shared decision-making for those receiving kidney replacement therapy, which has recently received a great deal of attention.Tables S2-

S6
Selection process 8 Specify the methods used to decide whether a study met the inclusion criteria of the review, including how many reviewers screened each record and each report retrieved, whether they worked independently, and if applicable, details of automation tools used in the process.

6-7
Data collection process 9 Specify the methods used to collect data from reports, including how many reviewers collected data from each report, whether they worked independently, any processes for obtaining or confirming data from study investigators, and if applicable, details of automation tools used in the process.

Data items 10a
List and define all outcomes for which data were sought.Specify whether all results that were compatible with each outcome domain in each study were sought (e.g. for all measures, time points, and analyses), and if not, the methods used to decide which results to collect.

10b
List and define all other variables for which data were sought (e.g.participant and intervention characteristics and funding sources).Describe any assumptions made about any missing or unclear information. 5 Study risk of bias assessment 11 Specify the methods used to assess risk of bias in the included studies, including details of the tool(s) used, how many reviewers assessed each study and whether they worked independently, and if applicable, details of automation tools used in the process.

7-8
Effect measures 12 Specify for each outcome the effect measure(s) (e.g.risk ratio and mean difference) used in the synthesis or presentation of results.

8-9
Synthesis methods 13a Describe the processes used to decide which studies were eligible for each synthesis (e.g.tabulating the study intervention characteristics and comparing against the planned groups for each synthesis (item #5)).

13b
Describe any methods required to prepare the data for presentation or synthesis, such as handling of missing summary statistics, or data conversions.9 13c Describe any methods used to tabulate or visually display results of individual studies and syntheses.9 Describe any methods used to synthesize results and provide a rationale for the choice(s).If meta-13d analysis was performed, describe the model(s), method(s) to identify the presence and extent of statistical heterogeneity, and software package(s) used.

13e
Describe any methods used to explore possible causes of heterogeneity among study results (e.g.subgroup analysis and meta-regression).9 13f Describe any sensitivity analyses conducted to assess robustness of the synthesized results.9 Reporting bias assessment 14 Describe any methods used to assess risk of bias due to missing results in a synthesis (arising from reporting biases).9 Certainty assessment 15 Describe any methods used to assess certainty (or confidence) in the body of evidence for an outcome.

Study selection 16a
Describe the results of the search and selection process, from the number of records identified in the search to the number of studies included in the review, ideally using a flow diagram.Figure 1 16b Cite studies that might appear to meet the inclusion criteria, but which were excluded, and explain why they were excluded.

FIGURE 2 :
FIGURE 2: One-year mortality after hip fracture in patients with ESKD ESKD, end-stage kidney disease

FIGURE 3 :
FIGURE 3: Five-year mortality after hip fracture in patients with ESKD ESKD, end-stage kidney disease

FIGURE 4 :FIGURE 5 :
FIGURE 4: One-year mortality after spinal fracture in patients with ESKD ESKD, end-stage kidney disease

FIGURE 6 :
FIGURE 6: Combined five-year mortality after hip and spinal fractures in patients with ESKD ESKD, end-stage kidney disease

Rationale 3 4 Objectives 4 4 METHODS Eligibility criteria 5 5 Search strategy 7
Describe the rationale for the review in the context of existing knowledge.3-Provide an explicit statement of the objective(s) or question(s) the review addresses.Specify the inclusion and exclusion criteria for the review and how studies were grouped for the syntheses., registers, websites, organisations, reference lists, and other sources searched or consulted to identify studies.Specify the date when each source was last searched or consulted.Present the full search strategies for all databases, registers, and websites, including any filters and limits used.

TABLE 1 : Summary of participant characteristics
a Standard deviation b Interquartile range c Range AU, Australia; CS, cohort study; CVD, cardiovascular diseases; DM, diabetes mellitus, EU, Europe; HD, hemodialysis; MC, multi centers; N/A, not available; NZ, New Zealand; PCT, prospective cohort study; PD, peritoneal dialysis; PY, person-year; RCT, retrospective cohort study; SC, single center

TABLE 3 : PRISMA 2020 checklist PRISMA
, Preferred Reporting Items for Systematic Reviews and Meta-Analyses