Dixon, Eleanor G.

Medical Research Council (MRC)

2025-12-04T12:50:34Z

2025-12-04T12:50:34Z

2025-12-04

Adverse drug reactions (ADRs) are defined as a noxious response to a medicinal product where a causal relationship between the medicinal product and an adverse event is either known or strongly suspected. ADRs account for 16.5% of UK hospital admissions, posing a growing global burden on healthcare and finances. Antibiotics are one of the most widely prescribed drug groups globally and are a major cause of ADRs; they are implicated in an estimated 19.3% of emergency visits relating to ADR events. Antibiotics are the leading cause of idiosyncratic drug-induced liver injury (DILI), causing an estimated 25-46% of all cases. Two leading causes of antibiotic-associated DILI are co-amoxiclav (in Europe) and anti-tuberculosis (TB) therapy (globally). DILI is the commonest cause of acute liver failure and has a mortality risk between 10-50%. Not only can ADRs cause patient mobility and mortality, they can cause treatment administration to pause to allow the patient to be treated for their ADR. In the case of antibiotics, this means the patient’s infection is not being treated, compromising treatment success. Current pharmacovigilance methods result in an estimated global under-reporting rate of ADRs of 94%. However, there is scope to address this through the use of routinely collected data (RCD)- data collected by organisations as part of their routine work. RCD include clinical data from electronic health records (EHR), disease registries and epidemiologic surveillance systems. Although RCD itself is not a new concept, the use of EHRs in pharmacoepidemiological research is expanding, catalysed through data linkage processes (the method of combining individual data sources using a unique identifier). Data linkage of EHRs results in combined patient-level data from different medical pathways and specialities. My thesis explores RCD’s utility in understanding antibiotic-associated ADRs by centring on two case studies exploring the role antibiotics play in DILI: a)UK and European Challenge: co-amoxiclav b)Global Challenge: anti-TB therapy. My thesis objectives are: 1.To undertake a proof-of-principle study investigating the ability of RCD to detect acute liver injury consistent with DILI in individuals treated with amoxicillin and co-amoxiclav (Chapter 2). 2.To identify knowledge gaps regarding the relationship between ADRs and missed doses of anti-TB therapy (Chapter 3). 3.To use RCD to determine the importance of ADRs as a driver of missed doses of anti-TB treatment (Chapter 4). For the first objective, I used routinely collected NHS records to identify cases of acute liver injury and DILI in patients prescribed amoxicillin or co-amoxiclav; a stepwise screening process guided by clinically used tools (e.g., the Roussel Uclaf Causality Assessment (RUCAM)) was used to identify patients with either condition (diagnosis of exclusion). I found that linked EHRs demonstrated a difference in the incidence risk of acute liver injury in patients prescribed co-amoxiclav (52.6 (CI: 46.2- 59.0) per 10,000 patients) compared to amoxicillin (24.1 (CI: 21.0- 27.3) per 10,000 patients). However, the difference in incidence risk of DILI between the two drug groups was minimal and inconsistent with the scarce literature values. For my second objective, I utilised a scoping review methodology to identify knowledge gaps regarding the relationship between missed doses of anti-TB therapy and ADRs. Patients of any age, with any site of disease and any drug susceptibility profile were included. ADRs were identified as a key driver of missed doses across patients in all drug susceptibility profiles yet information regarding the patterns of these missed doses was absent. Moreover, only 38% studies identified specific ADRs in relation to missed doses, confirming the underreporting of ADRs in literature and clinical practice. For my final objective, I used routinely collected medical notes and Directly Observed Therapy (DOT) records to address knowledge gaps identified in the scoping review. All patients were ≥18 years and were being treated for drug-sensitive TB using the same regimen. I confirmed that ADRs are a key driver of missed doses of anti-TB therapy and identified that ADRs cause longer yet less frequent periods of missed doses than missed doses for other reasons; 13.2% of doses missed due to ADRs for one dose in length compared to 56.4% for doses missed for other reasons. Hepatobiliary and gastrointestinal disorders caused the most doses of anti-TB therapy to be missed. Given the current interest and increasing use of RCD in pharmacovigilance activities, understanding the utility and limitations of RCD can inform future ADR monitoring efforts. Such knowledge facilitates a method of gaining current, real-world insights into patient-level exposure-outcome pathways, driving evidence-based policy and ultimately improving the safety of drugs.

https://hdl.handle.net/1842/44270

http://dx.doi.org/10.7488/era/6791

en

The University of Edinburgh

Dixon EG, Rasool S, Otaalo B, et al. No action is without its side effects: Adverse drug reactions and missed doses of antituberculosis therapy, a scoping review. British journal of clinical pharmacology 2024;90(1):313-20

2026-12-04

Adverse drug reactions (ADRs)

Antibiotics

Drug-induced liver injury (DILI)

Routinely collected data (RCD)

Pharmacoepidemiological research

Antibiotic-induced adverse drug reactions: investigating the utility of routinely collected data

Thesis or Dissertation

Doctoral

PhD Doctor of Philosophy

RESTRICTED ACCESS