Infected total hip arthroplasty - Bacteriology and the role of operating room ventilation in the reduction of postoperative infection

Abstract

Each year approximately 7,000 patients receive a total hip arthroplasty (THA) in Norway. The most common indication is osteoarthritis and most patients experience substantial pain relief postoperatively. A rare but much feared complication after arthroplasty surgery is postoperative prosthetic joint infection (PJI). This is a huge burden for patients and a technical challenge for surgeons, generating huge costs for the health care system and society. Knowledge of how and why these infections arise is essential in order to provide the best possible prophylaxis against and treatment of PJI. The aim of the first part of this thesis was to assess what kind of bacteria cause revisions due to infection after THA (Paper I). We also wanted to assess correlations between bacteriology and haematological findings (Paper I) and antibiotic resistance amongst the bacteria found in THA revisions due to infection (Paper II). We used the Norwegian Arthroplasty Register (NAR) and supplemented this with bacteriological findings including antibiotic resistance patterns and preoperative blood samples including parameters of infection, collected from patient records. As expected, we found a large proportion of Staphylococci, a commensal organism known to be part of normal human skin flora. Coagulase negative staphylococci (CoNS) showed increased resistance to several antibiotics that are used both as prophylaxis and in empirical and definitive treatment of PJI. As the staphylococcus aureus (S. aureus) species led to significantly higher biochemical parameters of infection compared to CoNS, our results might inform the choice of empirical treatment based on haematological status in cases of arthroplasty infection with an unknown causative pathogen. In the second part of the thesis (Papers III and IV), we wanted to enhance understanding of the origin of these infections by assessing operating room ventilation as a prophylactic measure against revision due to infection. The air in the operating room is, in addition to other surgery- and patient-related factors, a possible risk factor for postoperative infection. In Norwegian hospitals there are mainly two ventilation principles that are used to increase the cleanliness of the air: conventional, turbulent, mixing ventilation (CV) and unidirectional airflow (UDF) ventilation, formerly known as laminar airflow (LAF) ventilation, where the latter has in recent years been recommended on a disputed scientific basis. Lack of evidence for reduction in postoperative infection has led to the implementation of UDF systems being questioned as a prophylactic measure against postoperative infection. The NAR holds surgeon-reported data on the type of ventilation used during primary THA. We validated these ventilation data by performing a comprehensive assessment of the historical and present ventilation systems in 40 hospitals in Norway during the period 1987-2015 (Paper III). This was done in cooperation with knowledgeable surgeons and engineers at the relevant hospitals. This assessment showed that not all surgeons knew exactly what kind of ventilation system they performed the THA in, and accordingly may have reported inaccurate data to the register. This might have led to erroneous conclusions in earlier register studies on this topic. A series of such studies using data on ventilation reported by the surgeon or surgical department contributed to the scientific basis for a report on infection reduction from the World Health Organization (WHO) from 2016. This report concluded that UDF systems should not be installed in new operating rooms where arthroplasty is performed. We continued the project by conducting new analyses on the risk of revision due to infection after THA using validated ventilation data (Paper IV). This included sub- analyses of technical specifications of the different systems. We concluded that UDF systems do not increase the risk of infection, as recent literature seems to imply. By performing a sub-classification based on technical specifications, we show that there is substantial variation between the different UDF systems and that the more modern and large UDF systems, offering high volumes of air, show a slight reduction in the risk of revision due to infection after THA compared to CV. This is in concordance with other studies showing that UDF/LAF systems are able to create cleaner air than CV systems. When taking our results into account, considering also the finding of increased antibiotic resistance amongst common causative bacteria of THA, it would be erroneous to discontinue the use of large, high volume, vertical UDF systems in the operating room of the future. We hope and believe that the results of this thesis will have an impact on the ongoing international discussion on operating room ventilation and perioperative care.