Abstract
Aims
The Peri-Implant and PeriProsthetic Survival AnalysiS (PIPPAS) study aimed to investigate the risk factors for one-year mortality of femoral peri-implant fractures (FPIFs).
Methods
This prospective, multicentre, observational study involved 440 FPIF patients with a minimum one-year follow-up. Data on demographics, clinical features, fracture characteristics, management, and mortality rates were collected and analyzed using both univariate and multivariate analyses. FPIF patients were elderly (median age 87 years (IQR 81 to 92)), mostly female (82.5%, n = 363), and frail: median clinical frailty scale 6 (IQR 4 to 7), median Pfeiffer 4 (1 to 7), median age-adjusted Charlson Comorbidity Index (CCI) 6 (IQR 5 to 7), and 58.9% (n = 250) were American Society of Anesthesiologists grade III.
Results
Overall, 90.5% (n = 398) of the patients were treated surgically, 57.0% (n = 227) retained the implant, and 88.7% (n = 353) managed with fixation. Mortality rates were 8.2% (n = 3.6) in-hospital, 11.4% (n = 50) at 30 days, 21.1% (n = 93) at six months, and 21.6% (n = 95) at 12 months. Medical complications, mainly delirium, were common in the acute setting (52.7%, n = 215). The nonunion rate was 4.1% (n = 18). Mortality risk factors in the univariate analysis were age, living at a nursing home, no walking outdoors, frailty variables, fractures in the distal epiphysis, fractures around a proximal nail, discharge to a healthcare facility, and no osteoporotic treatment at discharge. Protective factors against mortality in the univariate analysis were surgical treatment by an experienced surgeon, management without an arthroplasty, allowing full weightbearing, mobilization in the first 48 hours postoperatively, and geriatric involvement. Risk factors for mortality in the multivariate analysis were cognitive impairment (Pfeiffer’s questionnaire) (hazard ratio (HR) 1.14 (95% CI 1.05 to 1.23), p = 0.002), age-adjusted CCI (HR 1.18 (95% CI 1.07 to 1.30), p = 0.001), and antiaggregant or anticoagulant medication at admission (HR 2.00 (95% CI 1.19 to 3.38), p = 0.009). Haemoglobin level at admission was protective against mortality (HR 0.85 (95% CI 0.74 to 0.97), p = 0.018).
Conclusion
Mortality in FPIFs occurs mainly within the first six months of follow-up. Early co-management and clinical optimization, particularly targeting frail older patients, is crucial in reducing mortality following these fractures.
Cite this article: Bone Jt Open 2024;6(1):43–52.
Take home message
Mortality rate associated with femoral peri-implant fractures (FPIFs) was 21% and predominantly occurred within the first six months of follow-up.
Risk factors for mortality in FPIFs were cognitive impairment, age-adjusted Charlson Comorbidity Index, and antiaggregant/anticoagulant medication at admission.
Early geriatric co-management and clinical optimization strategies can enhance survival outcomes in FPIF patients.
Introduction
Based on demographic trends and the increasing incidence of proximal femoral fractures, the prevalence of femoral peri-implant fractures (FPIFs) is expected to increase.1-3 This increase is linked to the growing population of individuals with non-prosthetic fixation devices, whose susceptibility to recurrent falls and new fractures grows with increasing age and comorbidities.3,4 We defined PIFs as a fracture occurring in bones with an existing non-prosthetic fixation device, such as plates, intramedullary nails, or screws.4-9
FPIFs, mainly involving older patients with multiple medical comorbidities,10 are associated with severe medical complications, prolonged hospital stays, and delayed recovery.4,11 The femoral periprosthetic fracture (FPPF) population shows mortality rates equal to or higher than those observed in the broader proximal femur fracture population,1,10,12-19 although FPIFs are distinct from FPPFs, and should be understood as a separate entity.
Challenges in the surgical treatment for FPIFs arise from factors related to the implant, the bone healing status, anatomical changes, and osteoporosis.2,5,20,21 Surgical treatment options and proposed algorithms are based on a relatively small number of patients.5,6,21 Most studies on FPIFs refer to cephalomedullary nails (CMNs), thus excluding diaphyseal and distal femoral implants.1,3,4,6,10,20-23 It is unknown whether different management methods affect mortality in FPIFs. Tools for decision-making are essential to reduce mortality and clinical complications, enhance functional outcomes, and preserve quality of life (QoL).
Given the rarity of FPIFs (0.5% to 2.3%),1,3,6,7,11,22-25 we conducted a multicentre, collaborative research project, Peri-Implant and Peri-Prosthetic fractures: AnalysiS (PIPPAS),15 to increase the number of patients included and provide more robust conclusions.1,5,6,10,11 This study aimed to evaluate the association between patient demographics, fracture characteristics, and fracture care on one-year mortality following FPIFs.
Methods
The PIPPAS study is a collaborative, multicentre, prospective observational case series study (level IV evidence) evaluating PPFs and PIFs in 56 Spanish hospitals and one in Argentina.15 FPIF management was the standard of care at each participating site, as determined by the attending surgeon. We included patients aged 18 years or older who presented between January 2021 and November 2022 with a FPIF following nail or plate fixation, and available one-year follow-up clinical data. Fractures between an implant and a stemless femoral component of a knee prosthesis were included only if the fracture fixation device was a greater determinant for the surgical management than the knee prosthesis; otherwise, fractures between an implant and a stem were considered a FPPF. We excluded patients with pathological or intraoperative fractures, failed fixation without a new fracture line, such as cut-out or cut-in, and pregnancy. Written consent for participation in the study was obtained from all participants or their legal representatives.
The index fracture refers to the fracture for which the implant related to the FPIF was originally used. Although several classifications have been used to describe FPIFs,5,6,21,26 many do not consider implants in the distal femur. We adopted the Broggi Classification for FPIFs,26 adding the healing status of the index fracture. We excluded Broggi-type D FPIFs from this analysis as they behave as a regular transverse subtrochanteric fracture, requiring screw removal (Figure 1).
Fig. 1
Diagram demonstrating the Broggi classification system for femoral peri-implant fractures, excluding type D fractures.
Prospective data collection included patient demographics, management, and outcomes based on the Fragility Fracture Network’s Minimum Common Dataset for hip fracture audits but adapted to the specific nature of FPIFs (Supplementary Material).27 Cognitive status was assessed with the Pfeiffer Short Portable Mental Status Questionnaire (SPMSQ).28 Health-related QoL was assessed using the EuroQol five-dimension five-level questionnaire (EQ-5D-5L) instrument at six and 12 months.29 Experienced surgeons were those who had performed over 20 minimally invasive fixations or arthroplasty revisions in the last 12 months. Fracture healing was defined as the presence of at least three cortical callus bridges on radiological examination and pain-free full weightbearing. A comprehensive list of variables is available in the Supplementary Material.
Data were collected and managed using REDCap electronic data capture tools hosted at Instituto de Estudio de Ciencias de la Salud de Castilla y León in Spain.30 The manuscript was adapted to the STrengthening the Reporting of OBservational studies in Epidemiology (STROBE) statement. The study was conducted in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki,31 and received approval from the institutional review boards of the coordinating centre and each participating hospital. This study is registered at ClinicalTrials.gov (NCT04663893).
The study involved 461 patients, of whom 440 met the inclusion criteria and 21 were lost to follow-up. The demographic and clinical characteristics of the participants are presented in Table I. Most patients were female (72.9%, n = 379), with a median age of 87 years (IQR 81 to 92). Most were frail, with a median Clinical Frailty Scale (CFS)32 of 6 (IQR 4 to 7), and mild cognitive impairment (median Pfeiffer SPMSQ 4 (IQR 1 to 7)). A large proportion were community-dwellers (72.5%, n = 317) and capable of walking outdoors (52.7%, n = 232).
Table I.
Demographic and baseline data for patients presenting a femoral peri-implant fracture.
| Variable | Value |
|---|---|
| Total patients, n | n = 440 |
| Median age, yrs (IQR) | 87 (81 to 92) |
| Sex, n (%) | |
| Female | 363 (82.5) |
| Male | 77 (17.5) |
| Place of residence, n (%) | |
| Own home | 317 (72.5) |
| Nursing home | 115 (26.3) |
| Hospital | 5 (1.1) |
| N/A | 3 (0.7) |
| Pre-fracture mobility,* n (%) | |
| 1 | 63 (14.4) |
| 2 | 73 (16.7) |
| 3 | 96 (21.9) |
| 4 | 111 (25.3) |
| 5 | 95 (21.7) |
| N/A | 2 (0.5) |
| Pfeiffer’s SPMSQ | |
| Median (IQR) | 4 (1 to 7) |
| N/A, n (%) | 27 (6.1) |
| CFS | |
| Median (IQR) | 6 (4 to 7) |
| N/A, n (%) | 10 (2.3) |
| ASA grade, n (%) | |
| I | 6 (1.6) |
| II | 81 (19.9) |
| III | 250 (58.9) |
| IV | 80 (19.2) |
| V | 2 (0.5) |
| N/A | 21 (4.8) |
| Median CCI (IQR) | 6 (5 to 7) |
| Osteoprotective treatment, n (%) | |
| No treatment | 250 (56.8) |
| Osteoprotective treatment | 190 (43.2) |
| Anti-resorptive | 64 (14.5) |
| Bone-forming | 8 (1.8) |
| Calcium | 121 (27.5) |
| Vitamin D | 150 (34.1) |
| Antiaggregant or anticoagulant medication, n (%) | |
| None | 289 (65.7) |
| Acenocumarol, NOAC, or PAA | 145 (33) |
| Double | 6 (1.4) |
| Hb at admission | |
| Median, gr/dL (IQR) | 11.9 (10.6 to 13.1) |
| N/A, n (%) | 3 (0.7) |
-
*
Pre-fracture mobility scale: 1 = complete independent gait; 2 = outdoors independent gait with one technical aid; 3 = outdoors independent gait with two technical aids; 4 = only indoors independent gait with or without aids; 5 = no mobility at all, or with the help of two other people.
-
ASA, American Society of Anesthesiologists; CCI, Charlson Comorbidity Index; CFS, clinical frailty scale; Hb, haemoglobin; N/A, not available; NOAC, new oral anticoagulant; PAA, platelet anti-aggregant; Pfeiffer’s SPMSQ, Pfeiffer’s Short Portable Mental Status Questionnaire.
Statistical analysis
Quantitative variables were summarized as medians and IQRs, and qualitative variables were presented according to their frequency distribution. Multivariate Cox regression analyses were performed to identify independent risk factors for one-year mortality after FPIF, including variables that reached a significance level of 0.1 in previous univariate analyses. A chi-squared test was used to compare observed results with expected results. The hazard ratio (HR) indicates the increase of the mortality risk for each unit a quantitative variable increases. If the HR < 1 the variable is a protective factor. Kaplan-Meier estimators were used to estimate survival functions, and comparisons were made using the log-rank test. A p-value < 0.05 were considered statistically significant. Statistical analyses were performed using SPSS v.29 software (IBM, USA).
Results
Patient mortality rates were as follows: 8.2% (n = 36) in-hospital, 11.4% (n = 50) within 30 days, 21.1% (n = 93) within six months, and 21.6% (n = 95) within one year, as illustrated by Kaplan-Meier curves in Figure 2. The incidence rate of FPIFs during 2021 was 2.3/105 individuals and 2.29% after femoral fracture fixation.
Fig. 2
Kaplan-Meier curves for one-year mortality. a) Global mortality curve (95% CI 9.710 to 10.168). b) Mortality curves for femoral peri-implant fracture (FPIF) around a nail (95% CI 9.053 to 9.999) and FPIF around a plate (95% CI 10.251 to 11.558) mortality curves. c) Mortality curves for management strategies “implant removed” (95% CI 10.133 to 11.164) vs “implant retained” (95% CI 9.326 to 10.410).
FPIF characteristics are detailed in Table II. The index fracture was already healed in 79.2% of cases. The type of FPIF according to Broggi’s classification is shown in Figure 3, with a greater risk of mortality for FPIFs around a nail than involving a plate (p = 0.007, univariate Cox regression analysis).
Table II.
Femoral peri-implant fracture diagnostic features.
| Variable | N (%) |
|---|---|
| ABC type | |
| A, at the tip of implant (epiphysis) | 73 (16.6) |
| B, at the tip of implant (diaphysis) | 127 (28.9) |
| C, distal to the tip of implant | 193 (43.9) |
| E, through the implant | 47 (10.7) |
| Implant | |
| Nail | 345 (78.4) |
| Nail proximal | 280 (63.6) |
| Nail distal | 65 (14.8) |
| Plate | 95 (21.6) |
| Plate proximal | 39 (8.9) |
| Plate distal | 56 (12.7) |
| FPIF bone segment location | |
| Proximal epiphysis | 113 (25.7) |
| Diaphysis | 227 (51.6) |
| Distal epiphysis | 100 (22.7) |
| Index fracture healed | |
| No | 90 (20.8) |
| Yes | 343 (79.2) |
| Previous infection | |
| No | 432 (98.2) |
| Yes | 8 (1.8) |
| Previous implant loosening radiological signs | |
| No | 423 (96.1) |
| Yes | 17 (3.9) |
| Implant displacement | |
| No | 395 (89.8) |
| Yes | 45 (10.2) |
| Time between index fracture to FPIF | |
| < 1 mth | 23 (5.4) |
| 1 to < 6 mths | 80 (18.8) |
| ≥ 6 to < 12 mths | 36 (8.5) |
| 1 to < 5 yrs | 114 (26.8) |
| ≥ 5 yrs | 172 (40.5) |
| N/A | 15 (3.4) |
-
FPIF, femoral peri-implant fracture; N/A, not available.
Fig. 3
Sankey diagrams (made in Flourish; Canva, Australia) with femoral peri-implant fracture (FPIF) type distribution according to Broggi’s classification. a) Bone segment where the FPIF is located related to FPIF subtypes according to the implant. b) A, B, C, and E types related to FPIF subtypes according to the implant.
Management strategies are outlined in Table III. Most patients (90.5%, n = 398) were treated surgically, primarily under spinal anaesthesia (89.1%, n = 334) after a median delay of 85.8 hours (IQR 47.0 to 132.6). Less invasive surgical approaches were used in 48.7% (n = 194) of patients and primary implants were removed in 42.5% (n = 169) of the cases. Multiple fixation techniques were used, most frequently one single plate (45.7%, n = 182), and included overlapping techniques to prevent stress risers (57%, n = 227). For patients not managed with a prosthesis, the use of a cerclage for reduction, open approaches, or retaining the previous implant were associated with unrestricted postoperative weightbearing (all p < 0.001), with no influence on the one-year mortality (p = 0.327, p = 0.931, p = 0.054, respectively; chi-squared test).
Table III.
Management of femoral peri-implant fractures. Categorical variables are summarized as absolute frequency and percentages from the number of patients managed surgically, except for “treatment”.
| FPIF | N = 440 |
|---|---|
| Treatment, n (%) | |
| Operative | 398 (90.5) |
| Nonoperative | 42 (9.5) |
| Surgical delay, hrs | |
| Median (IQR) | 85.8 (47.0 to 132.6) |
| N/A, n (%) | 2 (0.5) |
| Type of anaesthesia, n (%) * | |
| General | 77 (17.5) |
| Spinal/regional | 334 (89.1) |
| Surgical approach, n (%) | |
| Open | 202 (50.8) |
| MIS | 114 (28.6) |
| PC | 80 (20.1) |
| N/A | 2 (0.5) |
| Removal of previous implant, n (%) | |
| No | 227 (57.0) |
| Yes | 169 (42.5) |
| N/A | 2 (0.5) |
| Cerclage for reduction, n (%) | |
| No | 271 (68.1) |
| Yes | 125 (31.4) |
| N/A | 2 (0.5) |
| Arthroplasty, n (%) | |
| No | 353 (88.7) |
| Yes* | 43 (10.8) |
| N/A | 2 (0.5) |
| Type of fixation, n (%) † | 378 (95.0) |
| 1 plate | 182 (45.7) |
| 2 plates | 10 (2.5) |
| Nail | 171 (43) |
| Ex fix | 1 (0.3) |
| Cerclage | 68 (17.1) |
| Isolated screws | 9 (2.3) |
| Overlapping, mm | |
| Yes, n (%) | 227 (57.0) |
| Median (IQR) | 109 (61 to 154) |
| Gap, mm | |
| Yes, n (%) | 17 (4.3) |
| Median (IQR) | 25 (10 to 67) |
| Kissing implants, n (%) | 4 (1) |
| Interlocking, n (%) | |
| No | 315 (79.1) |
| Yes | 74 (18.6) |
| N/A | 9 (2.3) |
| Surgeon experience, n (%) | |
| > 20 arthroplasties | 73 (16.6) |
| > 20 MIPO | 153 (34.8) |
| Medical staff involved in the patient’s care, n (%) ‡ | |
| No | 75 (17) |
| Geriatrician | 134 (30.5) |
| Internal Medicine | 151 (34.3) |
| Geriatrician and others | 51 (11.6) |
| Others | 29 (6.6) |
| Initial postoperative mobilization out of bed, n (%) | |
| < 24 hrs | 118 (27.8) |
| 24 to 48 hrs | 194 (45.8) |
| > 48 hrs | 112 (26.4) |
| Median total length of hospital stay, hrs (IQR) | 266.0 (184.1 to 386.1) |
-
*
Patients under spinal anaesthesia could also receive general anaesthesia. An arthroplasty was used as part of the FPIF treatment, but it was not recorded whether cement was used or not.
-
†
The fixation strategy could include two or more fixation devices.
-
‡
Other than trauma or anaesthesia.
-
Ex fix, external fixator; FPIF, femoral peri-implant fracture; MIPO, minimally invasive plating ostheosynthesis; MIS, minimally invasive surgery; N/A, not available; PC, percutaneous.
Complications and secondary outcomes other than mortality are outlined in Table IV. Most fractures healed, with a nonunion rate of 4.1% (n = 18) at one year; only five patients (1.1%) were surgically treated for nonunion. An improvement in QoL was noted between six- and 12-month follow-up. Medical complications were commonly present in the acute episode, particularly delirium. After hospital discharge, pulmonary complications were the most common. Surgical complications mainly involved dislocations and prosthetic loosening among patients treated with arthroplasties, while patients managed with fixation had complications related to fracture healing and fixation failure. However, there were no differences in overall complications between both groups.
Table IV.
Postoperative care and follow-up data for femoral peri-implant fractures. Categorical variables are summarized as absolute frequency and percentages from the number of patients in each category.
| Medical complications | In-hospital | 30 days | 6 mths | 12 mths |
|---|---|---|---|---|
| No | 194 (47.4) | 326 (81.5) | 233 (71.5) | 209 (80.4) |
| Yes (any) | 215 (52.6) | 74 (18.5) | 93 (28.5) | 51 (19.6) |
| Cardiac | 59 (27.4) | 13 (3.3) | 20 (6.1) | 13 (5) |
| Pulmonary | 52 (24.2) | 20 (5) | 27 (8.3) | 23 (8.8) |
| Pulmonary thromboembolism | 3 (1.4) | 0 (0) | 1 (0.3) | 0 (0) |
| Renal | 65 (30.2) | 9 (2.3) | 8 (2.5) | 5 (1.9) |
| Cerebral | 6 (2.8) | 8 (2) | 7 (2.2) | 2 (0.8) |
| Gastrointestinal | 43 (20) | 12 (3) | 12 (3.7) | 7 (2.7) |
| Urinary tract infection | 47 (21.9) | 13 (3.3) | 10 (3.1) | 7 (2.7) |
| Delirium | 97 (45.1) | |||
| In-hospital fractures | 5 (2.3) | |||
| Other medical complications | 10 (2.5) | 25 (7.7) | 21 (8.1) | |
| Surgical complications | ||||
| No | 377 (94.3) | 296 (90.8) | 244 (93.8) | |
| Fracture in the same bone | 2 (0.5) | 7 (2.1) | 4 (1.5) | |
| Failure of fixation | 1 (0.3) | 1 (0.3) | 1 (0.4) | |
| Dislocation (prosthesis) | 16 (4) | 14 (4.3) | 5 (1.9) | |
| Loosen prosthesis | 7 (1.8) | 7 (2.1) | 2 (0.8) | |
| Infection | 0 (0) | 0 (0) | 0 (0) | |
| Nonunion | 5 (1.9) | |||
| Weightbearing restrictions | Hospital discharge | |||
| No restrictions | 165 (40.3) | 176 (45.1) | ||
| Only for transferences | 55 (13.4) | 75 (19.2) | ||
| Not allowed | 176 (43) | 137 (35.1) | ||
| N/A | 13 (3.2) | 2 (0.5) | ||
| Osteoprotective treatment* | ||||
| No treatment | 157 (38.4) | 167 (43.0) | 123 (42.3) | 97 (40.4) |
| Osteoprotective treatment | 252 (61.6) | 221 (57.0) | 168 (57.7) | 143 (59.6) |
| Antiresorptive | 94 (37.3) | 93 (42.1) | 68 (40.5) | 51 (35.7) |
| Bone-forming | 24 (9.5) | 20 (9.0) | 15 (8.9) | 16 (11.2) |
| Calcium | 153 (60.7) | 147 (66.6) | 110 (65.5) | 92 (63.3) |
| Vitamin D | 180 (71.4) | 186 (84.2) | 142 (84.5) | 118 (82.5) |
| Place of residence | ||||
| Home | 210 (51.3) | 218 (56.0) | 194 (66.7) | 166 (69.5) |
| Nursing home | 151 (36.9) | 150 (38.6) | 91 (31.6) | 71 (29.7) |
| Hospital | 12 (2.9) | 21 (5.4) | 5 (1.7) | 2 (0.8) |
| EQ-5D | 0.608 (0.575 to 0.641) | 0.694 (0.661 to 0.727) |
-
*
The percentages for the different osteoprotective treatments refer to the total number of patients who were receiving treatment.
-
EQ-5D, EuroQol five-dimension questionnaire.
Univariate analysis identified that patients who died were older, more frail (higher CFS, ASA score, and CCI), non-community dwellers, dependent ambulators, and did not receive osteoporosis treatment. Fractures in the distal epiphysis, around a proximal nail (vs a distal nail or a distal plate), first mobilization more than 48 hours after the operation, weightbearing restrictions, not involving a geriatrician, discharge to a skilled care facility, and not treating osteoporosis at discharge were associated with mortality; while FPIFs around a plate, surgical treatment, or surgery done by an experienced surgeon were protective factors (Supplementary Table i). Multivariate analysis showed that cognitive impairment, CCI, and preoperative antiaggregant or anticoagulant treatment were risk factors for mortality during the first-year post-fracture, whereas the haemoglobin level at admission was protective (Table V).
Table V.
Cox’s regression multivariate analysis.
| Variable | p-value | HR | 95% CI for HR |
|---|---|---|---|
| Removal of previous implant: yes vs no | 0.228 | 0.723 | 0.427 to 1.225 |
| Cognitive impairment (Pfeiffer’s SPMSQ) | 0.002 | 1.135 | 1.048 to 1.229 |
| CCI (age-adjusted) | 0.001 | 1.178 | 1.067 to 1.301 |
| Haemoglobin at admission, g/dl | 0.018 | 0.850 | 0.743 to 0.973 |
| Antiaggregant or anticoagulant medication at admission (no ref) | 0.033 | ||
| Either Acenocumarol or NOAC or PAA | 0.009 | 2.000 | 1.185 to 3.376 |
| Double | 0.414 | 1.861 | 0.419 to 8.253 |
-
CCI, Charlson Comorbidity Index; HR, hazard ratio; NOAC, new oral anticoagulant; PAA, platelet antiaggregant; Pfeiffer’s SPMSQ, Pfeiffer’s Short Portable Mental Status Questionnaire.
Discussion
To the best of our knowledge, this is the largest study on FPIFs assessing risk factors for one-year mortality. Despite their rarity, FPIFs are a severe complication in this elderly and frail population.3 The working hypothesis in most studies on FPIFs is whether short and long CMN have similar FPIF rates, resulting in a small reported number of FPIFs. Previous studies have reported incidences between 1.4% and 2.0%,1,24 with a slightly higher rate of 2.3% observed in Spain, likely due to the country’s higher life expectancy. The rate of FPIFs around CMN is decreasing with newer generations of nails.25
Compared to other studies on FPIFs, patients in this study were older, with a greater proportion of females,1,3-6,21,24,33 likely attributable to higher life expectancy in Spain. Information on patients’ comorbidities or fracture baseline data is scarce; in the study by Jennison and Yarlagadda10 patients were aged 12 years younger, but with worse ASA grades. Frailty, limited mobility (no outdoor ambulation), and living in nursing homes were candidate predictors of mortality in the univariate analysis. Like the PIPPAS study, previous studies report that 70% to 80% of index fractures had already healed at the time the FPIF occurred.1,5,7,11,22 Only Bidolegui et al21 and Lindvall et al23 presented all index fractures healed. Typically, FPIFs occurred more than one year after the index fracture fixation,3,4,6,7,22,34 except in rare cases where they happened after 1.5 months.24,25
The nail-to-plate ratio in this study was 4:1, but this ratio varies across literature from 1:1 to 1:2.5,5,6,21 influenced by the treatment chosen for the index fracture. Some authors only report FPIFs on nails and exclude plates.7,22,33 As in most series, the most involved segment was the diaphysis, with fractures located at the tip of the implant.1,5,6,21,23,33 Distal FPIFs, particularly type C fractures occurring distal to a proximal nail or in the distal metaphysis, were associated with increased mortality. As in this study, Müller et al1 found that mortality was higher for nail-related FPIFs.
One-year mortality rates vary across studies, ranging from 18.6% to 44.8%.1,3,7,10,11 As in Lang et al,11 patients mainly died within the first six months. Jennison and Yarlagadda10 reported a double mortality rate, with 23% of 29 patients deceased at one-month follow-up. These differences can be explained by the heterogeneity of the studies. FPIF mortality is higher than that in hip fracture patients; the latter has benefited from practices to improve outcomes including national hip fracture audits.10 PIPPAS aims to raise awareness that patients with FPIFs would also benefit from these measures.
Some authors failed to find significant risk factors for mortality in patients with FPIF.10,11 Prompt surgical treatment, and lower CCI10 and ASA grades,11 have shown a trend towards survival. This study confirmed that reduced general health status, as evidenced by higher CCI, cognitive impairment or antiaggregant/anticoagulant medication, and lower haemoglobin levels at admission, are risk factors for mortality. Optimization of health status upon arrival, including anaemia, can help reduce mortality. Surgical management is usually discarded in patients deemed unfit for surgery. Clinical optimization and prompt management of medical complications can be better achieved with geriatric co-management.
Strategies that are protective candidate predictors should be considered during fracture care: surgical treatment by an experienced surgeon, removing the implant, with a new fixation allowing full weightbearing and early postoperative mobilization while co-managed with a geriatrician. The reason for removing the previous implant was mostly to use intramedullary devices, i.e. a nail or a prosthesis. Clinically, it is unclear whether or not removing the implant is better for patient survival, especially if different appropriate treatment alternatives exist that do not require implant removal. For example, current treatment options for diaphyseal FPIFs distal to CMN could be either adding a plate or changing the nail for a longer nail. Goodnough et al33 found that revision to a longer nail for FPIFs around a CMN in 16 patients was associated with increased patient morbidity. Our study, by contrast, found a trend towards removing the implant as a protective candidate predictor against mortality, but included FPIFs around plates and nails. It is important to consider that if the index fracture remains unhealed, and the patient then suffers a second fracture (the FPIF), then there will be two fracture sites in the same bone. Both fracture sites need adequate fixation in order to heal, although the healing status of the index fracture had no influence on mortality.
Patients who were not mobilized promptly after surgery, or without unrestricted weightbearing, were also more prone to be discharged to skilled care facilities. Medical or surgical complications at any time were risk factors for mortality. Post-fracture osteoporosis treatment protected against mortality in the univariate analysis; such treatment could be a confounding factor, as poly-medicated patients with more comorbidities are less likely to start treatment for osteoporosis. Furthermore, patients treated by geriatricians, who are more prone to initiate osteoporosis treatment, were more likely to survive.
Several classification systems for PIFs have been proposed, but none have gained widespread acceptance.3-6 The original FPIF classification by Videla-Cés et al26 has been validated, with a high inter- and intraobserver correlation rate. Treatment algorithms based on FPIF classifications are unable to compare outcomes of different operative strategies4,5 and encourage future multicentre research to study expected outcomes.6 The PISCO group presented a consensus review on FPIF treatment with expert recommendations, but it is yet to be determined whether they reduce mortality and complication rates.9
The aims of treatment are ‘getting it right first time’ with a single operation, which allows immediate unrestricted weightbearing with a low risk of complications, and avoids the creation of stress risers locally which may predispose to further PIFs, using a long implant to span the entire femur, or at least overlap the original implant.2,5 There is a wide range of surgical strategies,4,7,11,24 with little information on mortality and functional outcomes regarding walking ability and place of residence. Lang et al11 recommended the use of cerclage wires as a reduction tool through an open approach. However, to avoid potential harm, we recommend less invasive approaches. Poroh et al4 reported a nonunion rate of 4.2%, similar to the 4.1% observed in the present study, while in other studies all FPIFs healed.1,5,21,33
The present study found an association of demographics, fracture characteristics, and treatment strategy with one-year mortality following FPIFs. However, a detailed analysis of the treatment options for each fracture type, including implant and fracture location, could provide more well-defined recommendations on surgical management.
Limitations of this study include, first, its heterogeneous nature in terms of implants used, FPIF locations, and healing status of the index fractures, which complicates the formulation of conclusive general guidelines. Second, there were no monitoring visits of the participating sites, and no external confirmation to limit the number of unconscious errors each site may have made when entering data in REDCap. Lastly, a more detailed analysis of specific FPIF patterns could give more precise guidance for surgical treatment strategy. Despite these limitations, a comprehensive understanding of these fractures can assist readers in organizing their resources.
In conclusion, mortality associated with FPIF predominantly occurred within the first six months of follow-up. Risk factors for mortality were cognitive impairment, age-adjusted CCI, and antiaggregant/anticoagulant medication at admission. Conversely, higher admission haemoglobin levels proved to be protective against mortality. Implementing early geriatric co-management and clinical optimization strategies, especially for frail patients, can enhance survival outcomes in FPIF cases.
References
1. Müller F , Galler M , Zellner M , Bäuml C , Marzouk A , Füchtmeier B . Peri-implant femoral fractures: the risk is more than three times higher within PFN compared with DHS . Injury . 2016 ; 47 ( 10 ): 2189 – 2194 . Crossref PubMed Google Scholar
2. Liporace FA , Yoon RS , Collinge CA . Interprosthetic and peri-implant fractures: principles of operative fixation and future directions . J Orthop Trauma . 2017 ; 31 ( 5 ): 287 – 292 . Crossref PubMed Google Scholar
3. Kruse M , Mohammed J , Sayed-Noor A , Wolf O , Holmgren G , Nordström R , et al. Peri-implant femoral fractures in hip fracture patients treated with osteosynthesis: a retrospective cohort study of 1965 patients . Eur J Trauma Emerg Surg . 2022 ; 48 ( 1 ): 293 – 298 . Crossref PubMed Google Scholar
4. Poroh M , Puha B , Gheorghevici TS , et al. A retrospective analysis of peri-implant fractures: insights from a large volume clinical study . Int Orthop . 2023 ; 47 ( 11 ): 2859 – 2868 . Crossref PubMed Google Scholar
5. Chan LWM , Gardner AW , Wong MK , Chua K , Kwek EBK , Kagda F , et al. Non-prosthetic peri-implant fractures: classification, management and outcomes . Arch Orthop Trauma Surg . 2018 ; 138 ( 6 ): 791 – 802 . Crossref PubMed Google Scholar
6. Egol KA , Carlock KD , Kelly EA , et al. Previous implant fractures: a new descriptive classification system . J Orthop Trauma . 2019 ; 33 ( 9 ): 423 – 427 . Crossref PubMed Google Scholar
7. Vilar-Sastre I , Corró S , Tomàs-Hernández J , et al. Fractures after cephalomedullary nailing of the femur: systematization of surgical fixation based on the analysis of a single-center retrospective cohort . Int Orthop . 2022 ; 46 ( 10 ): 2357 – 2364 . Crossref PubMed Google Scholar
8. Perskin CR , Seetharam A , Mullis BH , et al. Peri-implant fractures of the upper and lower extremities: a case series of 61 fractures . Eur J Orthop Surg Traumatol . 2022 ; 32 ( 3 ): 467 – 474 . Crossref PubMed Google Scholar
9. Castillón P , Muñoz Vives JM , Aguado HJ , et al. Consensus review on peri-implant femur fracture treatment: Peri-Implant Spanish Consensus (PISCO) investigators’ recommendations . EFORT Open Rev . 2024 ; 9 ( 1 ): 40 – 50 . Crossref Google Scholar
10. Jennison T , Yarlagadda R . Mortality in patients sustaining a periprosthetic fracture following a previous extracapsular hip fracture fixation . Injury . 2018 ; 49 ( 3 ): 702 – 704 . Crossref PubMed Google Scholar
11. Lang NW , Joestl J , Payr S , Platzer P , Sarahrudi K . Secondary femur shaft fracture following treatment with cephalomedullary nail: a retrospective single-center experience . Arch Orthop Trauma Surg . 2017 ; 137 ( 9 ): 1271 – 1278 . Crossref PubMed Google Scholar
12. Nasser A , Prakash R , Handford C , et al. Predictors of mortality in periprosthetic fractures of the hip: results from the national PPF study . Injury . 2023 ; 54 ( 12 ). Crossref PubMed Google Scholar
13. Gitajn IL , Heng M , Weaver MJ , et al. Mortality following surgical management of vancouver B periprosthetic fractures . J Orthop Trauma . 2017 ; 31 ( 1 ): 9 – 14 . Crossref PubMed Google Scholar
14. Halonen LM , Stenroos A , Vasara H , Kosola J . Peri-implant fracture: a rare complication after intramedullary fixation of trochanteric femoral fracture . Arch Orthop Trauma Surg . 2022 ; 142 ( 12 ): 3715 – 3720 . Crossref PubMed Google Scholar
15. Aguado HJ , Castillón-Bernal P , Teixidor-Serra J , García-Sánchez Y , Muñoz-Vives JM , Camacho-Carrasco P , et al. Optimizing periprosthetic fracture management and in-hospital outcome: insights from the PIPPAS multicentric study of 1387 cases in spain . J Orthop Traumatol . 2024 ; 25 ( 1 ). Crossref PubMed Google Scholar
16. Kjærvik C , Gjertsen JE , Stensland E , Saltyte-Benth J , Soereide O . Modifiable and non-modifiable risk factors in hip fracture mortality in norway, 2014 to 2018: a linked multiregistry study . Bone Joint J . 2022 ; 104-B ( 7 ): 884 – 893 . PubMed Google Scholar
17. Jain S , Farook MZ , Aslam-Pervez N , et al. A multicentre comparative analysis of fixation versus revision surgery for periprosthetic femoral fractures following total hip arthroplasty with a cemented polished taper-slip femoral component . Bone Joint J . 2023 ; 105 ( B(2 ): 124 – 134 . Crossref PubMed Google Scholar
18. Baker P , Kottam L , Walker R , Dwyer K , Jameson S , Scrimshire A . Epidemiology and characteristics of femoral periprosthetic fractures . Bone Joint J . 2022 ; 104-B ( 8 ): 987 – 996 . Crossref PubMed Google Scholar
19. Szymski D , Walter N , Krull P , et al. Infection after intracapsular femoral neck fracture - does antibiotic-loaded bone cement reduce infection risk after hemiarthroplasty and total hip arthroplasty? Bone Joint Res . 2023 ; 12 ( 5 ): 331 – 338 . Crossref PubMed Google Scholar
20. Yamamoto N , Yamakawa Y , Inokuchi T , et al. Hip fractures following intramedullary nailing fixation for femoral fractures . Injury . 2022 ; 53 ( 3 ): 1190 – 1195 . Crossref PubMed Google Scholar
21. Bidolegui F , Pereira S , Munera MA , et al. Peri-implant femoral fractures: challenges, outcomes, and proposal of a treatment algorithm . Chin J Traumatol . 2023 ; 26 ( 4 ): 211 – 216 . Crossref PubMed Google Scholar
22. Larose G , Tufescu T , Graham C . Periprosthetic fracture rate after short and long hip nails: analysis of a regional health database . Injury . 2022 ; 53 ( 6 ): 2195 – 2198 . PubMed Crossref Google Scholar
23. Lindvall E , Ghaffar S , Martirosian A , Husak L . Short versus long intramedullary nails in the treatment of pertrochanteric hip fractures: incidence of ipsilateral fractures and costs associated with each implant . J Orthop Trauma . 2016 ; 30 ( 3 ): 119 – 124 . Crossref PubMed Google Scholar
24. Skála-Rosenbaum J , Džupa V , Bartoška R , Douša P , Waldauf P , Krbec M . Distal locking in short hip nails: cause or prevention of peri-implant fractures? Injury . 2016 ; 47 ( 4 ): 887 – 892 . Crossref PubMed Google Scholar
25. Norris R , Bhattacharjee D , Parker MJ . Occurrence of secondary fracture around intramedullary nails used for trochanteric hip fractures: a systematic review of 13,568 patients . Injury . 2012 ; 43 ( 6 ): 706 – 711 . Crossref PubMed Google Scholar
26. Videla-Cés M , Sales-Pérez J-M , Sánchez-Navés R , Romero-Pijoan E , Videla S . Proposal for the classification of peri-implant femoral fractures: retrospective cohort study . Injury . 2019 ; 50 ( 3 ): 758 – 763 . Crossref PubMed Google Scholar
27. Fragility Fracture Network . Hip Fracture Audit - Minimum Common Dataset . https://fragilityfracturenetwork.org/hip-fracture-audit/ ( date last accessed 19 November 2024 ). Google Scholar
28. Pfeiffer E . A short portable mental status questionnaire for the assessment of organic brain deficit in elderly patients . J Am Geriatr Soc . 1975 ; 23 ( 10 ): 433 – 441 . Crossref PubMed Google Scholar
29. Herdman M , Gudex C , Lloyd A , et al. Development and preliminary testing of the new five-level version of EQ-5D (EQ-5D-5L) . Qual Life Res . 2011 ; 20 ( 10 ): 1727 – 1736 . Crossref PubMed Google Scholar
30. Harris PA , Taylor R , Minor BL , et al. The REDCap consortium: building an international community of software platform partners . J Biomed Inform . 2019 ; 95 . Crossref PubMed Google Scholar
31. World Medical Association . World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects . JAMA . 2013 ; 310 ( 20 ): 2191 – 2194 . Crossref PubMed Google Scholar
32. Rockwood K , Song X , MacKnight C , et al. A global clinical measure of fitness and frailty in elderly people . CMAJ . 2005 ; 173 ( 5 ): 489 – 495 . Crossref PubMed Google Scholar
33. Goodnough LH , Salazar BP , Furness J , et al. How are peri-implant fractures below short versus long cephalomedullary nails different? Eur J Orthop Surg Traumatol . 2021 ; 31 ( 3 ): 421 – 427 . Crossref PubMed Google Scholar
34. Kleweno C , Morgan J , Redshaw J , et al. Short versus long cephalomedullary nails for the treatment of intertrochanteric hip fractures in patients older than 65 years . J Orthop Trauma . 2014 ; 28 ( 7 ): 391 – 397 . Crossref PubMed Google Scholar
Author contributions
H. J. Aguado: Writing – original draft, Data curation, Investigation, Software, Validation, Formal analysis, Methodology, Project administration, Writing – review & editing
Funding statement
The study was funded by grants from AO Trauma Foundation (Spain) (grant number: 20-2041) and SACYL (Sanidad de Castilla y León) (grant number: GRS: 2371/A/21). No funding entity had a role in the design or conduct of the study, the collection or analyses of data, or manuscript preparation.
ICMJE COI statement
H. J. Aguado reports funding from SACYL (GRS 2371/A/21) and the AO Foundation (AO Trauma Spain) (PI-20-2041).
Data sharing
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Acknowledgements
The Peri-Implant and PeriProsthetic Survival Analysis (PIPPAS) study was organized by seven Trauma and Orthopaedic Surgery departments in Spain, the PIPPAS coordinating team. HJA was the principal investigator and Hospital Clínico Universitario de Valladolid the lead centre and HJA, APM, EMA, SPO and VGV responsible for the database, data validation, analyses, and study-centre coordination. The Coordinating team (HJA, PCB, JTS, YGS, JMMV, PCC, MJG, COT, PGP and APM) designed the study and vouched for the accuracy of the data and analyses. JMM, RCG and HJA performed the statistical analysis. HJA wrote the first draft of the manuscript, and HJA, PCB, JTS, YGS, JMMV, PCC, MJG, COT, PGP and APM made critical revisions. All participating investigators in the PIPPAS Study Group made comments, read, approved the final manuscript, and decided to submit the paper.
We thank Dasha Gorbenko del Blanco (PhD, medical illustrator, dasha.blanco@gmail.com) for her artwork producing the figures and infographics for the PIPPAS study.
The PIPPAS Study Group: Héctor J Aguado, Pablo Castillón-Bernal, Jordi Teixidor-Serra, Yaiza García-Sánchez, Josep M. Muñoz-Vives, Pilar Camacho-Carrasco, Cristina Ojeda-Thies, Pablo García-Portabella, Adela Pereda-Manso, Elvira Mateos-Álvarez, Virginia García-Virto, David Noriega-González, Begoña Aránzazu Álvarez-Ramos, Mª Fe Muñoz-Moreno, Irene Arroyo-Hernantes, Carmen Martínez-Sellés, Sergio Marín-Jiménez, Adriana Acha, Jordi Tomás-Hernández, Jordi Selga-Marsà, José Vicente Andrés-Peiró, Carlos Piedra-Calle, Ferrán Blasco-Casado, Ernesto Guerra-Farfán, Jordi Querolt-Coll, Guillermo Triana-López de Santamaría, Carles Gil-Aliberas, Borja Campuzano-Bitterling, Eliam Ajuria Fernández, Rebeca Díaz Suárez, Eugenia Fernández Manzano, Unai García De Cortázar, Mirentxu Arrieta, Daniel Escobar, Estíbaliz Castrillo MD, Patricia Balvis Balvis, Maciej Denisiuk, Jesús Moreta, Xabier Uriarte, Andrea Vea, Patricia Jiménez-Tellería, Beatriz Olías-López, Patricia Amaya-Espinosa, Juan Boluda-Mengod MD, Juan Carlos Borrás-Cebrián, Carles Martínez-Pérez, Patricio Andrés Freile Pazmiño, Pablo Calavia-Calé, Miguel Ángel Suárez-Suárez, Antonio García Arias, José Valle-Cruz, Javier García-Coiradas, María Ángeles Cano Leira, César Bonome-Roel, Antonio Benjumea Carrasco, Marcos Chico-García, Coral Sánchez Pérez, Rodrigo Jesús Priego Sánchez, Ana López-de Pariza, Alexis Fernández-Juan, Eladio Saura-Sánchez, Sandra Giménez-Ibáñez, Plácido Sánchez-Gómez, F. Javier Ricón-Recarey, Jesús Jiménez-Olivares, Silvia Correoso-Castellanos, Elena M. García García, Isabel Medrano-Morte, Francisco Cuadrado-Abajo, María Esther Laguna-Bercero, Pedro del Pozo-Manrique, Francisco Manuel García Navas-García, Ester García-Paredero, Teresa Beteta Robles, Inés Navas-Pernía, Gonzalo Gálvez-Márquez, Ignasi De Villasante-Jirón, Joan Vilanova-Laguna, Miquel Videla-Ces, Teresa Serra Porta, Gloria González-Ojeda, Carmen Carrasco Becerra, Silvia Pena Paz, Fátima Fernández-Dorado, Amaia Martínez-Menduiña, Víctor Vaquerizo-García, Antonio Murcia-Asensio, Elena Galián-Muñoz, Carmelo Marín-Martínez, Adrián Muñoz-Vicente, Nuria Plaza-Salazar, Carla Gámez-Asunción, Jennifer Benito-Santamaría, Ana Verdejo González, Laura Alonso Viana, Juan Mingo-Robinet, Amaya Barbería-Biurrun, Emma Escudero-Martínez, Laura Chouza-Montero, María Naharro-Tobío, Alfons Gasset-Teixidor, Andrea Domínguez-Ibarrola, J M Peñalver, Jorge Serrano-Sanz, Adrián Roche-Albero, Carlos Martín-Hernández, María Macho-Mier, Julián Carlos Segura-Nuez, José Carlos Saló-Cuenca, Jordi Espona Roselló, Guillermo Criado-Albillos, Hugo Gabriel Cabello-Benavides, David Alonso Nestar, Jorge Martínez-Íñiguez Blasco, José Manuel Bogallo-Dorado, Juan Ramón Cano-Porras, Fernando Marqués-López, Santos Martínez-Díaz, Guido S. Carabelli, Pablo I. Slullitel, Ignacio Astore, Carlos Hernández-Pascual, Javier Marín-Sánchez, Iván Dot-Pascuet, Ana Piñeiro-Borrero, José María Pérez-Sánchez, Alfonso Mandía-Martínez, Julio De Caso-Rodríguez, Jordi Martín-Marcuello, Miguel Benito-Mateo, Ainhoa Jaúregui-Garasa, María Rosa González-Panisello, Marta Miñana-Barrios, Susana Iglesias-Fernández, Raquel García-Albea, María Carmen González-López.
Ethical review statement
All patients or their relatives gave their written consent to be included in the study. The study coordinating centre and each participating hospital obtained institutional review board approval. The study was approved by the coordinating centre ethics committee board: “Comité Ético de Investigación Médica Área de Salud Valladolid Este” code: PI 20-2041. The “Comité Ético de Investigación Médica Área de Salud Valladolid Este” is the reference ethics committee board for the PIPPAS Study.
Open access funding
The open access fee for this article was provided by SACYL (Sanidad de Castilla y León) (grant number: GRS 2794/A1/2023).
Trial registration number
Registered at ClinicalTrials.gov (NCT04663893) protocol ID: PI 20-2041.
Supplementary material
Univariate analysis table, and data collected from patients presenting with a peri-implant fracture.
© 2025 Aguado et al. Open Access This article is distributed under the terms of the Creative Commons Attributions (CC BY 4.0) licence (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium or format, provided the original author and source are credited.
