Responsible antibiotic use and diagnostic challenges in infectious diseases : studies in a resource-limited setting and a high-income setting

Abstract

Background: Antibiotic resistance is a globally emerging health challenge. In resource-limited settings, the burdens of infectious diseases and of antibiotic resistance are the highest. The World Health Organization (WHO) has stated that high and uncontrolled use of antibiotics in resource-limited settings contributes to the increasing prevalence of antibiotic resistance in such areas. Improvements should be based on knowledge about current antibiotic use, preventing the spread of infections, and the implementation and/or development of diagnostic methods for the improved management of infectious diseases. Limited access to diagnostic methods complicates the management of infectious diseases in resource-limited settings and increases the risk of disregarding severe infectious diseases, such as infective endocarditis (IE). Therefore, it is important to evaluate diagnostic methods that can improve the management of infectious diseases in such settings. Evaluating whether the findings of echocardiography (ECHO) could give a suggestion of the bacterial aetiology in patients with IE in a high-income setting might provide valuable information for the implementation of ECHO in resource-limited settings.

Aims: To map and describe the antibiotic prescribing practices and to highlight areas of improvement in the management of in-patients with infectious diseases in a resource-limited setting in India. Further to assess whether specific manifestations detected by ECHO were associated with certain bacterial species in patients with IE in a high-income setting in Sweden.

Methods: The studies for Papers I-III were conducted in a teaching hospital (TH) and a non-teaching (NTH) hospital in India. Two were prospective, cross-sectional studies (I and II) and one was a time series analysis (III). Paper IV was conducted in Stockholm, Sweden, as a register-based cohort study. For Paper I, antibiotic use was analysed at the medical intensive care units (ICUs) at the two hospitals in India. For Paper II, antibiotic prescribing practices were analysed with a focus on infectious diagnoses at the paediatric departments at the two hospitals in India. For Paper III, time series analyses over a 10-year period (2008–2017) were conducted to follow antibiotic prescribing over time among patients with severe infections at the NTH and the TH in India. For Paper IV, associations between IE manifestations detected by ECHO and bacterial species were assessed from a cohort of patients with IE obtained from the Swedish National Registry of Infective Endocarditis (SRIE).

Main findings: High percentages of patients at the medical ICUs were prescribed antibiotics at both hospitals (4,843 of 6,141 patients, 70%) although less than 25% of the patients were registered with an infection-associated diagnosis (I). At the paediatric departments, antibiotic use among patients with either acute gastroenteritis, respiratory tract infections, enteric fever, viral fever or unspecified fever was more common at the NTH (2,088 patients, 84%) compared with the TH (224 patients, 44%; P < 0.001). Broad-spectrum antibiotics were commonly prescribed at both hospitals, and less than 40% of the prescribed antibiotics at both hospitals were compliant to the national list of essential medicines (37% at the TH and 24% at the NTH; P < 0.05) (II). From 2008 to 2017, the overall rate of antibiotic prescribing for patients with severe infections, as well as prescribing of first- and second-line antibiotics and so-called fixed-dose combinations (FDCs) of antibiotics increased in the NTH (P < 0.05). In the TH, the overall antibiotic prescribing did not change significantly, although the prescribing of second- and third-line antibiotics and FDCs of antibiotics increased during the study period (P < 0.05) (III). Among patients with IE, associations were seen between aortic valve vegetation and Enterococcus faecalis among patients with native aortic valves, between mitral valve vegetation and streptococci of group B or viridans group streptococci, between tricuspid valve vegetation and Staphylococcus aureus among patients with intravenous drug abuse, and between perivalvular abscesses as well as cardiovascular implantable electronic device (CIED)-associated IE and coagulase-negative staphylococci (all P < 0.05) (IV).

Conclusions: Antibiotic prescribing was high at the ICUs and paediatric departments at the NTH and the TH, also among patients presenting with no infection-associated diagnoses or with diagnoses for which antibiotic treatments are not recommended. Broad-spectrum antibiotics were commonly prescribed at both hospitals. From 2008 to 2017, prescribing of second-line antibiotics and FDCs of antibiotics among patients with severe infections increased at both hospitals. The findings suggest that there is a need for improved antibiotic prescribing and management of infectious diseases. A major problem in this setting as well as in similar resource-limited settings is the limited use of diagnostics, which are needed to guide the treatment and follow-up of infectious diseases. Limited use of diagnostics also increases the risk of underdiagnosing infectious diseases. Very few cases of IE were registered at the NTH and TH, which might be a result of the limited use of diagnostics. ECHO is an effective method for the diagnosis, evaluation and follow-up of patients with IE, successfully implemented in high-income settings. ECHO can be used to identify cardiac manifestations of IE but might also give a suggestion of bacterial aetiology as some IE manifestations detected by ECHO have been shown to be associated with certain bacterial species. These findings could possibly be useful also in resource-limited settings, where ECHO might improve the management of patients with IE.