Aims

To develop and validate patient-centred algorithms that estimate individual risk of death over the first year after elective joint arthroplasty surgery for osteoarthritis.

The development of an algorithm that provides accurate individualised estimates of revision risk could help patients make informed surgical treatment choices. This requires building a survival model based on fixed and modifiable risk factors that predict outcome at the individual level. Here we compare different survival models for predicting prosthesis survivorship after hip replacement for osteoarthritis using data from the National Joint Registry for England, Wales, Northern Ireland and the Isle of Man.

In this comparative study we implemented parametric and flexible parametric (FP) methods and random survival forests (RSF). The overall performance of the parametric models was compared using Akaike information criterion (AIC). The preferred parametric model and the RSF algorithm were further compared in terms of the Brier score, concordance index (C index) and calibration.

The dataset contains 327 238 hip replacements for osteoarthritis carried out in England and Wales between 2003 and 2015. The AIC value for the FP model was the lowest. The averages of survival probability estimates were in good agreement with the observed values for the FP model and the RSF algorithm. The integrated Brier score of the FP model and the RSF approach over 10 years were similar: 0.011 (95% confidence interval: 0.011–0.011). The C index of the FP model at 10 years was 59.4% (95% confidence interval: 59.4%–59.4%). This was 56.2% (56.1%–56.3%) for the RSF method.

The FP model outperformed other commonly used survival models across chosen validation criteria. However, it does not provide high discriminatory power at the individual level. Models with more comprehensive risk adjustment may provide additional insights for individual risk.

Introduction

The development of an algorithm that provides accurate individualised estimates of revision risk could help patients make informed surgical treatment choices. This requires building a survival model based on fixed and modifiable risk factors that predict outcome at the individual level. Here we compare different survival models for predicting prosthesis survivorship after hip replacement for osteoarthritis using data from the National Joint Registry for England, Wales, Northern Ireland and the Isle of Man (NJR).

Background

Total knee replacement (TKR) is an effective operation for many patients, however approximately 20% of patients experience chronic pain and functional limitations in the months and years following their TKR. If modifiable pre-operative risk factors could be identified, this would allow patients to be targeted with individualised care to optimise these factors prior to surgery and potentially improve outcomes. Psychosocial factors have also been found to be important in predicting outcomes in the first 12 months after TKR, however their impact on long-term outcomes is unknown. This study aimed to identify pre-operative psychosocial predictors of patient-reported and clinician-assessed outcomes at one year and five years after primary TKR.

Aim

The aim of this study was to gain insight into the in vivo expression of virulence and metabolic genes of Staphylococcus aureus in a prosthetic joint infection in a human subject.

Culture of multiple intraoperative tissue samples is the standard of microbiological diagnosis of prosthetic joint infections. Recently, improved sensitivity of using prosthesis sonication method and molecular techniques has been reported in the literature. However, collecting the removed prosthesis as well as additional specimens for molecular analysis is not straightforward for the surgeons and assistants in the operation theatre. Our All-in-a-Box concept addresses the need for simple and unambiguous sampling of clinical specimens in the operating theatre, and to overcome the variation in sampling technique within and between surgical teams and across different hospitals.

The All-in-a-Box concept was developed in close cooperation between surgeons, their operating assistants, clinical microbiologists and molecular biologists in order to ensure the concept is easily implemented in the operating theatre, achieving high completeness, and being well preserved all the way to the laboratory.

All needed equipment, vials and forms are collected in a single box, and corresponding items are clearly color coded to further reduce the likelihood of confusion.

Boxes are designed to address the specific needs for either routine diagnosis or special demands as in clinical studies. Their design is based on large experience in connection with diagnosis of joint prosthesis-related infections. Downstream SOPs for sample processing are included in the All-in-a-box concept and specimens can subsequently be analyzed in parallel by culturing and molecular methods.

We have implemented this concept in two large research projects, we received 1508 (89%) of 1685 scheduled samples during the 2-year project period in the first project despite several different surgical teams and hospitals, while the other project is still ongoing.

All-in-a-Box is useful concept to improve the completeness of routine sampling for microbial analysis.

Identification of modalities and procedures to improve the differential diagnosis of septic and aseptic cases in patients with joint-related pain after total hip or knee alloplasty (THA/TKA).

A prospective cohort of 147 patients presenting with problems related to previous THA or TKA was included and subjected to a comprehensive diagnostic algorithm. The standard diagnostics were supplemented with novel or improved methods for sampling of clinical specimens, sonication of retrieved implant parts, prolonged and effective culture of microorganisms, and dedicated clinical samples for molecular biological detection and identification of microorganisms. Furthermore, comprehensive pain investigations and nuclear imaging were employed. For each case the clinical management was decided upon in a clinical conference with participation of clinical microbiologist, orthopedics and experts in nuclear imaging. The clinical management of patients was blinded against the molecular biological detection of microorganisms.

Patients grouped as follows: 69 aseptic, 19 acute septic, 19 chronic septic, 40 pain/unresolved. Sonication of retrieved implant parts resulted in detection of biofilm not detected by standard specimens, i.e. joint fluid and periprosthetic tissue biopsies. Next generation sequencing detected and identified few infections not detected by culture. Molecular analyses showed more polymicrobial infections than culture. Nuclear imaging was inconclusive with respect to recommendation of changed setup. Analysis of blood based biomarkers is ongoing. Patients with chronic pain are undergoing follow-up.

The special emphasis put on detection of infections resulted in detection of infections in joints that otherwise would have been categorized as aseptic loosening. Clinical management for these cases was changed accordingly. The cross-disciplinary clinical conference is considered valuable for clinical management. The clinical relevance of the polymicrobial nature of infections as diagnosed employing next generation sequencing is yet to be established. Long-term follow-up is planned.

Recent evidence suggests that the microbial community, its spatial distribution and activity play an important role in the prolongation of treatment and healing of chronic infections. Standard bacterial cultures often underestimate the microbial diversity present in chronic infections. This lack of growth is often due to a combination of inadequate growth conditions, prior usage of antibiotics and presence of slow-growing, fastidious, anaerobic or unculturable bacteria living in biofilms. Thus, diagnosis of chronic infections is challenged by lack of appropriate sampling strategies and by limitations in microbiological testing methods. The purpose of this study was to improve sampling and diagnosis of prosthetic joint infections (PJI) and chronic wounds, especially considering the biofilm issue.

Systematic sampling, sonication of prosthesis and extended culture were applied on patients with chronic wounds and patients with suspected PJIs. Optimized DNA extraction, quantitative PCR, cloning, next generation sequencing and PNA FISH were applied on the different types of specimens for optimized diagnosis. For further investigation of the microbial pathogenesis, in situ transcriptomics and metabolomics were applied.

In both chronic wounds and PJIs, molecular techniques detected a larger diversity of microorganisms than culture methods in several patients. Especially in wounds, molecular methods identified more anaerobic pathogens than culture methods. A heterogeneous distribution of bacteria in various specimens from the same patient was evident for both patient groups. In chronic wounds, multiple biopsies from the same ulcer showed large differences in the abundance of S. aureus and P. aeruginosa at different locations. Transcriptomic and metabolomic analyses indicated the important virulence genes and nutrient acquisition mechanisms of Staphylococcus aureus in situ. As an example, diagnosis and treatment of a patient with a chronic biofilm prosthesis infection persisting for 7 years will be presented.

Our studies show that diagnosis of chronic biofilm related infections required multiple specimen types, standardized sampling, extended culture and molecular analysis. Our results are useful for improvement of sampling, analysis and treatment in the clinic. It is our ambition to translate studies on bacterial activity into clinical practice in the future.

Optimal sampling for culture-based or molecular diagnosis remains highly contested for patients suspected of prosthetic joint infection (PJI). Most existing studies have a retrospective design without a standardized sampling strategy. Therefore, the results are difficult to translate into guidelines. We have conducted a 2-year prospective study with a sampling strategy adaptable to the specific procedure in patients with either hip or knee alloplasty. Thus, comparisons of results obtained with different specimen types and diagnostic methods are possible.

The study enrolled patients with a painful hip or knee alloplasty. The sampling strategy for microbiological diagnosis included multiple specimens of each type (joint fluid, tissue biopsies, bone biopsies, and swabs taken from the prosthesis in situ), and prosthetic components (if removed). Prepacked boxes with containers and accessories for sampling, transport and storage were provided. Microbial culture and bacterial 16S rDNA screening were carried out for all specimen types. Whenever positive upon 16S rDNA screening, samples were analyzed further by sequencing. Peptide nucleic acid-fluorescence in situ hybridization (optimized using filtrations; Filter-PNA-FISH) was limited to a subset thereof.

An overall completeness of ∼90% was obtained by the sampling strategy in 164 procedures (‘cases’) in 131 patients. In 58 cases PJI was suspected, and a revision was carried out. 42 cases were culture-positive, and 16 were culture-negative; one culture-negative case was positive by 16S rDNA sequencing of a corresponding specimen. The contribution to a microbiological diagnosis was high for periprosthetic tissue biopsies (≥ 3 positive out of 5) 90%, prosthetic component(s) 90%, and joint fluid 94%. Conversely, the contribution was sparse for prosthetic swabs 50% and bone biopsies 40%, respectively. Filter-PNA-FISH was used to confirm findings by culture and to demonstrate biofilm formation.

With the described sampling strategy we reached high completeness of complex specimen sets. The sampling strategy may be adapted to other clinical settings with microbiological sampling of similar complexity. We found multiple periprosthetic tissue biopsies, prosthetic component(s) and joint fluid to form the optimal specimen set for culture-based diagnosis. The contribution by 16S rDNA sequencing is still under investigations but the contributions seems moderate probably because of a low rate of antibiotic therapy before the procedure, use of effective culture methods and prolonged incubation (14 days).