Development and Internal Validation of a Multivariable Prediction Model for Mortality After Hip Fracture with Machine Learning Techniques

In order to estimate the likelihood of 1, 3, 6 and 12 month mortality in patients with hip fractures, we applied a variety of machine learning methods using readily available, preoperative data. We used prospectively collected data from a single university hospital in Copenhagen, Denmark for consecutive patients with hip fractures, aged 60 years and older, treated between September 2008 to September 2010 (n = 1186). Preoperative biochemical and anamnestic data were used as predictors and outcome was survival at 1, 3, 6 and 12 months after the fracture. After feature selection for each timepoint a stratified split was done (70/30) before training and validating Random Forest models, extreme gradient boosting (XGB) and Generalized Linear Models. We evaluated and compared each model using receiver operator characteristic (ROC), calibration slope and intercept, Spiegelhalter's z- test and Decision Curve Analysis. Using combinations of between 10 and 13 anamnestic and biochemical parameters we were able to successfully estimate the likelihood of mortality with an area under the curve on ROC curves of 0.79, 0.80, 0.79 and 0.81 for 1, 3, 6 and 12 month, respectively. The XGB was the overall best calibrated and most promising model. The XGB model most successfully estimated the likelihood of mortality postoperatively. An easy-to-use model could be helpful in perioperative decisions concerning level of care, focused research and information to patients. External validation is necessary before widespread use and is currently underway, an online tool has been developed for educational/experimental purposes ( https://hipfx.shinyapps.io/hipfx/ ).

Diagnosis of Open Tibial Fracture Showed High Positive Predictive Value in the Swedish National Patient Register

Introduction

The Swedish National Patient Register was validated only for a few diagnoses in the field of trauma. In this study, we calculated the positive predictive values (PPV) of the diagnosis of open tibial fracture and corresponding E-codes (cause of injury).

Patients and Methods

Out of 2845 cases from a 10-year period (2007–2016), a random sample of 300 cases was selected for review of medical records. The accuracy of the diagnosis and cause of injury was calculated and presented as PPV. We divided the study population into two subgroups (moderate and severe injury) that were analyzed separately. Severe injury was defined as when a patient had an amputation and/or reconstructive surgical procedures, indicated by corresponding ICD-codes.

Results

The PPV of the diagnosis of open tibial fracture was 87% (95% CI: 86–88%) overall, 86% (95% CI: 79–91%) for moderate injuries and 96% (95% CI: 91–98%) for severe injuries. The PPV for E-codes was 74% (95% CI: 65–81%). The majority of injuries were caused by falls (47%) or transport accidents (38%). Most of these injuries were caused by high-energy trauma (60%).

Conclusion

The PPV of the diagnosis of open tibial fracture in the Swedish National Patient Register is high (87%). The PPV of E-codes was lower (79%). The results imply that the register is well suited for healthcare evaluation and research purposes regarding trauma diagnoses. Most open tibial fractures are high-energy injuries.

[Patient blood management - to transfuse blood on appropriate indications]

During the last decade, the varying use and the lack of consistent indications for blood transfusions have been questioned. Comparisons of liberal and restrictive transfusion policies, most often support a restrictive policy. This has led to an evidence-based approach to optimizing the care of patients who might need transfusion, Patient Blood Management (PBM). There is evidence that both anemia and allogeneic blood transfusions are independently associated with increased morbidity and mortality. In elective surgery it is possible to adapt the three pillars of PBM in a structured way; i.e. optimization of red blood cell mass, reduction of blood loss and bleeding, and optimization of the patient's physiological tolerance towards anemia. These activities should be included in the pre-peri- and postoperative routines, in all surgical units.

Hyperbaric Oxygen in Lower Limb Trauma (HOLLT); protocol for a randomised controlled trial

INTRODUCTION

Open fractures with significant soft tissue injury are associated with high rates of complications, such as non-union, infection, chronic pain and disability. Complications often require further inpatient care, and in many cases, multiple operations and prolonged rehabilitation. Use of hyperbaric oxygen therapy as an adjunct to standard orthopaedic trauma care has the potential to reduce the complications of musculoskeletal injury and thus improve outcomes. Two previous randomised trials have suggested some positive effect, but neither functional measures nor long-term outcomes were reported.

METHODS AND ANALYSIS

An international, multicentre, randomised, open-label, clinical trial. Patients with trauma with an acute open fracture of the tibia with severe soft tissue injury (Gustilo grade 3) and high risk of injury-related complications were recruited from participating major trauma hospitals with hyperbaric facilities. Patients were enrolled with the expectation of commencing 12 sessions of hyperbaric oxygen therapy within 48 h of injury. The primary outcome measure is the incidence of acute complications of the open fracture wound at 14 days. Other short-term outcome measures include amputation, need for fasciotomy, time until wound closure, breakdown of closed wounds, time until definitive orthopaedic fixation, number of operative procedures, intensive care stay and hospital stay. Long-term follow-up will continue for 2 years postinjury.

ETHICS AND DISSEMINATION

Ethics approval was given by The Alfred Health Human Ethics Committee (206/04) and the Monash University Human Research Ethics Committee (CF07/4208). Approval was also obtained from the institutional research ethics committee at each participating site. This study will make a significant contribution to the trauma literature and should answer the question of whether hyperbaric oxygen therapy can significantly improve outcomes in severe lower limb trauma. Collective study results will be published in international journals and presented at relevant conferences.

TRIAL REGISTRATION NUMBER

Clinicaltrials.gov: NCT00264511; Australian New Zealand Clinical Trials Registry (ANZCTR): ACTRN12607000559415.

Effects of a formalized collaboration between plastic and orthopedic surgeons in severe extremity trauma patients; a retrospective study

BACKGROUND

Severe trauma to the extremities often includes a combination of fractures and soft tissue injuries. Several publications support that the patient outcome is better when skeletal stabilization is followed by early soft-tissue coverage. In an effort to optimize the treatment of these patients, we established a formalized collaboration in 2008 between the Departments of reconstructive plastic surgery and orthopedics at the Karolinska University Hospital.

METHODS

A retrospective review was conducted for all patients who had suffered severe extremity trauma and received either a free or a pedicled flap for extremity reconstruction. We compared the management of patients 0-4 years before and 0-4 years after the collaboration started especially with respect to; choice of flap, time to flap coverage, number of operations/revisions, total in-hospital stay.

RESULTS

After initiation of the collaboration, the number of flaps increased from 13 flaps (5 free and 8 pedicled) to 44 flaps (21 free and 23 pedicled). Fewer postoperative revisions was seen, as well as shorter in-hospital stay.

CONCLUSIONS

The present study highlights the importance of formalized collaboration between orthopedic and plastic surgeons in severe extremity trauma patients. The concept of an interdisciplinary approach has led to an increased number of trauma patients referred for plastic surgical consultation, an increased number of flaps, fewer postoperative revisions and shorter hospital stay.

Lower extremity soft tissue reconstruction and amputation rates in patients with open tibial fractures in Sweden during 1998-2010

Background

The rates of soft tissue reconstruction and amputation after open tibial fractures have not been studied on a national perspective. We aimed to determine the frequency of soft tissue coverage after open tibial fracture as well as primary and secondary amputation rates.

Methods

Data on all patients (> = 15 years) admitted to hospital with open tibial fractures were extracted from the Swedish National Patient Register (1998–2010). All surgical procedures, re-admissions, and mechanisms of injury were analysed accordingly. The risk of amputation was calculated using logistic regression (adjusted for age, sex, mechanism of injury, reconstructive surgery and fixation method). The mean follow-up time was 6 (SD 3.8) years.

Results

Of 3,777 patients, 342 patients underwent soft tissue reconstructive surgery. In total, there were 125 amputations. Among patients with no reconstructive surgery, 2% (n = 68 patients) underwent amputation. In an adjusted analysis, patients older than 70 years (OR = 2.7, 95%, CI = 1.1-6) and those who underwent reconstructive surgery (OR = 3.1, 95% CI = 1.6-5.8) showed higher risk for amputation. Fixations other than intramedullary nailing (plate, external fixation, closed reduction and combination) as the only method were associated with a significant higher risk for amputation (OR 5.1-14.4). Reconstruction within 72 hours (3 days) showed better results than reconstruction between 4–90 days (p = 0.04).

Conclusions

The rate of amputations after open tibial fractures is low (3.6%). There is a higher risk for amputations with age above 70 (in contrast: male sex and tissue reconstruction are rather indicators for more severe soft tissue injuries). Only a small proportion of open tibial fractures need soft tissue reconstructive surgery. Reconstruction with free or pedicled flap should be performed within 72 hours whenever possible.

The value of preoperative grade of radiographic and histological changes in predicting pain relief after total knee arthroplasty for osteoarthritis

PURPOSE

Many attempts with contradictory results have been made to correlate different features of OA with pain. One reason may be that pain at rest and pain with movement are seldom considered separately although the mechanisms may be quite different. Furthermore, pain ratings are subject to individual interpretation making an inter-individual comparison questionable. By instead calculating the absolute and relative changes in pain on an intra-individual level after total knee arthroplasty (TKA), we aimed at exploring a relationship between pain and radiological and histological changes.

METHODS

In 69 patients undergoing TKA, preoperative radiographs and perioperative histological samples of the synovial membrane were graded for severity of osteoarthritic and inflammatory changes. The findings were related to the intensity of pain at rest and with movement both preoperatively and 18 months postoperatively according to the visual analogue scale (VAS).

RESULTS

The radiographic and histological findings showed no significant correlation with the mean pre- or postoperative pain scores. Instead, change in pain with movement from pre- to postoperative was significantly related to the grade of radiographic osteoarthritis. Best pain relief by TKA was achieved in patients with severe radiographic changes. This, however, only applied to pain with movement.

CONCLUSIONS

Pain at rest and pain with movement may have different mechanisms. We believe that assessing the intensity of pain at rest and pain with movement separately and considering changes in pain on an individual level will be helpful strategies in future follow-up studies and efforts aimed at explaining the mechanisms of pain in OA.

Epidemiology of adult ankle fractures in Sweden between 1987 and 2004: a population-based study of 91,410 Swedish inpatients

BACKGROUND AND PURPOSE

Previous national epidemiological data on the characteristics and trends of patients with ankle fractures have been limited. We therefore analyzed data on Swedish inpatients with ankle fractures in this nationwide population study, based on data from 1987 through 2004.

PATIENTS AND METHODS

Data on all inpatients aged 15 years and older with ankle fracture were extracted from the Swedish National Patient Register for the period 1987-2004.

RESULTS

We identified 91,410 hospital admissions with ankle fracture, corresponding to an annual incidence rate of 71 per 10(5) person-years. During the study period, the number of hospital admissions increased by 0.2% annually, mainly from increase in fracture incidence in the elderly women. Mean age at admission was 45 (SD 19) years for men and 58 (18) for women. The major mechanism of injury was falling at the same level (64%).

INTERPRETATION

This nationwide study of inpatients with ankle fractures showed an increase in fracture incidence, particularly in elderly women.

Health-related quality of life (EQ-5D) before and after orthopedic surgery

BACKGROUND AND PURPOSE

Population data on mortality and life expectancy are generally available for most countries. However, no longitudinal data based on the health-related quality of life outcome from the EQ-5D instrument have been reported for orthopedic patients. We assessed the effect of orthopedic surgery as measured by EQ-5D.

METHODS

We analyzed EQ-5D data from 2,444 patients who were operated at the Department of Orthopedic Surgery at Karolinska University Hospital, 2001-2005. We also made a comparison between results from this cohort and those from a Swedish EQ-5D population survey.

RESULTS

The mean EQ-5D (index) score improved from 0.54 to 0.72. Hip and knee arthroplasty, operations related to previous surgery, trauma-related procedures, and rheumatoid arthritis surgeries had preoperative EQ-5D (index) scores of 0.48 to 0.52. All of these groups showed substantial improvement in scores (0.63 to 0.80). Patients with tumors or diseases of the elbow/hand showed higher preoperative scores (0.66 to 0.77), which were similar postoperatively. In most patients, the EQ-5D (index) score improved but did not reach the level reported for an age- and sex-matched population sample (mean difference = 0.11).

INTERPRETATION

Our results can be used as part of the preoperative patient information to increase the level of patient awareness and cooperation, and to facilitate rehabilitation. In future it will be possible-but not easy-to use the EQ-5D instrument as a complementary consideration in clinical priority assessment.

Survival, complications and outcome in 282 patients operated for neurological deficit due to thoracic or lumbar spinal metastases

We present survival, neurological function, and complications in a consecutive series of 282 patients operated for spinal metastases from January 1990 to December 2001. Our main surgical indication throughout this time period was neurological deficit rather than pain. Metastases from cancer of the prostate accounted for 40%, breast 15%, kidney 8%, and lung 7%. In 78% the level of decompression was thoracic and lumbar in 22%. Thirteen percent had a single metastases only, 64% had multiple skeletal metastases, and 23% had non-skeletal metastases also. Preoperatively 64% were non-walkers (Frankel A-C), 30% could walk with aids (Frankel D) and 8% had normal motor function (Frankel E). Posterior decompression and stabilization was applied in 212 patients, 47 had laminectomy only, and 23 had anterior decompressions and reconstruction. Complications were recorded at a level of 20%, and systemic complications were often associated with early death. The survival rate was 0.63 at 3 months, 0.47 at 6 months, 0.30 at 1 year, and 0.16 at 2 years. Twelve of 255 (5%) patients with motor deficits were worsened postoperatively, whereas 179 (70%) improved at least one Frankel grade. The ability to walk postoperatively was retained during follow-up in more than 80% of the patients. This study shows that important improvement of function can be gained by surgical treatment, but the complication rate was high and many patients died of their disease within the first months of surgery.