Bloodstream infection (blood poisoning) and antibiotic resistance are increasing worldwide, and already cause the loss of millions of human lives each year. According to the World Health Organisation (WHO), bloodstream infections (BSIs) represent 20% of global mortality on a par with cardiac infarct, stroke, and major trauma. BSI may occur when bacteria from a focus of infection gain access to the circulation (bacteraemia). BSIs are usually divided into two subclasses: community- and hospital-onset infections, since disease this involves different patient groups, types of bacteria, and reasons for infection. Compared to other countries, Sweden has been fortunate in having a relatively low death rate from BSI and low antibiotic resistance. However, as our lifestyle changes, the age of the population increases with more disease as a result, and as the healthcare system responds, death from infection and antibiotic resistance are on the increase. It is important that we recognise ”warning symptoms” if we are to manage BSIs correctly and initiate effective treatment. It is difficult to design individualised empirical treatment, so it is very important to be aware of risk factors for BSI and local resistance patterns, and to have an effective management programme.
Bacterial resistance to antibiotics is an increasing problem, especially in bowel organisms that can cause infections that are very difficult to treat. In short, antibiotic resistance arises as a result of evolutionary processes where bacteria protect themselves by developing resistance genes. These genes can be exchanged between similar organisms or transmitted to others that in turn cause resistant infection. The use of antibiotics leads to an evolutionary/selection process leading to resistance in bacteria, both normal and pathogenic, enabling resistant organisms to survive, thrive, and go on to cause infection. Antibiotic resistance is a threat to global health.
This thesis aims to increase our awareness of a large group of patients who suffer bloodstream infection. BSIs are increasing globally, and the death toll is high. Antibiotic resistance is an increasing threat to the health of the population, and we are inundated by alarming reports of resistance getting out of control. What is the situation in Sweden, and can we identify risk factors for BSI and mortality?
In Study I, our aim was to study the incidence and mortality of BSI in Östergötland. To be able to do this, a large patient population stretching over several years was required. The study design was thus population-based in the form of an observational cohort study where all blood culture results from 2000 to 2013 were analysed, and evaluated from clinical data. A total of 109,938 results were analysed resulting in 11,480 BSIs.
We saw that the incidence of BSI increased by 64% (mostly community-onset BSIs). We also saw that mortality increased by 45%. These results illustrate the importance of nationwide cooperation to combat the increasing problem of BSI and its mortality, and the establishment of a nationwide BSI register.
The aim of Study II was to assess resistance development in Östergötland and its relationship to mortality. A total of 9,587 microorganisms were analysed between 2008 and 2016. We observed an increase in quinolone resistance (3.7-7.7%) and cephalosporin resistance (2.5-5.2%) amongst Enterobacteriaceae. We then looked at BSIs caused by multiresistant bacteria showing a total of 245 cases (2.6%); an increase of 300%. Despite this, we did not see an increased mortality in this group.
There are several possible explanations for the increase in BSI mortality of which antibiotic resistance is a predominant factor globally. We were unable to show this in our study, even so mortality is increasing and is currently at a high level. In Study III we therefore analysed risk factors associated with death during a community-acquired BSI, focusing on preliminary prehospital and hospital management. In a retrospective case-control study on 195 deaths matched 1:1 regarding age, gender, and microorganism, with 195 survivors (controls). Results showed that many patients had contacted the primary healthcare system because of infection before they became severely ill, and that the strongest affectable risk factor for death was delay (>24h) between primary healthcare visit and admission to hospital. This shows the need for increased awareness in society and amongst the medical profession of those patients at risk and symptoms that should raise the alarm, leading to more rapid treatment.
In Studies I and II we found an increase in both BSIs and mortality, we also saw an increase in antibiotic resistance and multiresistant bacteria, mainly ESBL-producing E. coli. On the other hand, we did not see any coupling between multiresistance and mortality in this Swedish population. E. coli is a gram-negative bacteria that causes most BSIs. Since E. coli is predominantly a urine tract pathogen, Study IV aimed to study BSIs caused by ESBL-producing E. coli originating from the urinary tract. We studied the prevalence of E. coli clones, resistance genes and risk factors, as well as any signs of increased mortality from ESBL-producing E. coli compared to sensitive E. coli. Our main finding was a surprisingly low mortality from ESBL-producing E. coli (3%). Most patients in the ESBL-producing E. coli group received inadequate antibiotic treatment for at least 48h, but we did not see any sign of increased mortality or risk for serious sepsis with circulatory failure in this group. This finding is interesting and opens up for new studies on virulence factors and immunological factors that govern the immune response to BSI.
The implementation of cost-effective monitoring systems including clinical microbiological epidemiology and early identification of BSI, together with information campaigns aimed at the public as well as healthcare personnel regarding patients at risk and symptoms giving cause for alarm, should lead to a radical reduction in morbidity and mortality from BSI. This requires new diagnostic tools to individualise both antibiotic treatment and targeted management based on microorganism virulence factors. Modernisation of the medical journal system with algorithms aimed at early identification of risk patients and automated suggestions for empirical antibiotic treatment based on antibiotic resistance seen in previous cultures and local resistance patterns, would certainly improve management. Furthermore, new immunological tests showing the type of immunological reaction to a serious BSI will lead to individualised immunotherapy that, together with antibiotic treatment, will further improve patient care in this important group.