Adverse drug reaction-related hospitalisation in Ethiopia

Adverse drug reactions (ADRs) are an important healthcare problem frequently associated with significant morbidity, hospitalisation and mortality. Globally, the prevalence of ADR-related hospitalisation and mortality vary from 0.2% to 54.5% and 0.1% to 10.0%, respectively. Severe ADRs are important reason for admission to intensive care unit and extensions of hospital stay in approximately one-fifth of overall ADR-related admissions. Overall, the rates of ADRrelated hospital admissions and mortality are comparable between developed and developing countries. However, there are marked differences between developed and developing countries with regard to the nature of the ADRs implicated in hospital admissions and the mortality rate. In addition, there are some important differences in risk factors contributing to ADR-related hospital admissions and mortality due to the differences in population socio-demographics, disease characteristics, drug therapy used, healthcare systems and ethnic origins. In Ethiopia, a developing country, there are number of factors thought to increase the risk of ADR-related hospital admission. These include, but are not limited to, a greater proportion of patients who take anti-tubercular (anti-TB) drugs and antiretroviral therapy (ART), a high prevalence of malnutrition and anaemia, a higher prevalence of concomitant anti-TB drugs and ART use, and widespread use of traditional remedies. In addition, there is a higher proportion of the slow acetylator phenotype among patients on ART and anti-TB drugs that increases susceptibility to ADRs. There is growing attention to chronic disease management with new and complex therapies in ambulatory care clinics, where there are higher rates of drug-related problems and irrational use of medicines that could lead to drug-related harm. Unlike developed countries, there is substantial all-cause mortality rate among patients presenting to emergency departments, a high rate of mortality among HIV/TB co-infected patients on drug therapy, a less health-literate population and a lesser ability to provide healthcare. Moreover, there are increasing rates of concurrent infectious and non-communicable diseases demanding multiple medications with potential drug interactions. The majority of studies focussing on risk factors for ADR-related hospitalisation and mortality have been conducted in developed countries. In Ethiopia, to our knowledge, there are no studies reporting on the prevalence and risk factors associated with ADR-related hospitalisation, or the mortality rate attributable to ADRs in patients presenting to hospital. The limited information available in the Ethiopian setting, the presence of multiple factors suspected to increase the risk of ADR-related admissions and evidence of a substantial burden of ADR-related admissions and mortality in other settings provided the impetus for this study. Determining the magnitude of ADR-related hospitalisation and mortality and identifying factors contributing to ADRs for community-based patients are crucial in understanding the extent of the problem and developing preventive strategies to decrease the clinical and economic burden. Therefore, the main aims of the body of work presented in this thesis were to identify ADRs responsible for ADR-related hospital admissions; investigate the medications and other risk factors associated with the ADRs; and to determine their severity, preventability, clinical presentation and outcomes. Due to the absence of similar studies in Ethiopian patients, we began by reviewing the existing literature on the prevalence and contributing factors of ADR-related hospitalisations in developed and developing countries. From 43 relevant publications identified through systematic review, the median (with interquartile range (IQR)) prevalence of ADR-related hospitalisations in developed and developing countries were 6.3% (3.3-11.0%) and 5.5% (1.1-16.9%), respectively. Similarly, the median proportions of ADR-related mortality in developed and developing countries were 1.7% (0.7-4.8%) and 1.8% (0.8-8.0%), respectively. Older age, female gender, number of medications, renal impairment and heart failure were reported to be associated with an increased risk for ADR-related hospitalisation in both settings, while HIV/AIDS was implicated in developing countries only. The majority of ADRs were preventable in both settings, highlighting the importance of improving medication use, particularly in vulnerable patient groups such as the elderly, patients with multiple comorbidities and, in developing countries, with HIV/AIDS. Following review of existing literature, a prospective observational study determined the prevalence of ADR-related hospitalisation, characterised the ADR types and their preventability, characterised the implicated medications and identified predictors of ADR related hospitalisation. This was determined through detailed review of medical records, laboratory tests and patient interviews followed by causality assessment by the Naranjo algorithm and expert consensus. Of 1,001 patients included, 103 (10.3%) were deemed to have experienced an ADR-related admission. Common ADRs responsible for hospitalisation were hepatotoxicity (35, 29.4%) followed by acute kidney injury (27, 22.7%) and electrolyte disturbances (hypokalaemia and hypocalcaemia) (13, 10.9%). The drug classes most frequently involved in ADRs were anti-TB drugs (36 patients, 35.0%), followed by ART (22 patients, 21.4%) and diuretics (19 patients, 18.4%). Body mass index (BMI) <18.5 kg/m2 (adjusted odd ratio [AOR]=1.69; 95% confidence interval [CI]=1.10-2.62), pre-existing renal disease (AOR=2.84; 95% CI=1.38-5.85), pre-existing liver disease (AOR=2.61; 95% CI=1.38-4.96), number of comorbidities ‚Äöv¢‚Ä¢4 (AOR=2.09; 95% CI=1.27-3.44), number of drugs ‚Äöv¢‚Ä¢6 (AOR=2.02; 95% CI=1.26-3.25) and history of previous ADRs (AOR=24.27; 95% CI=11.29-52.17) were found to be independent predictors of ADR-related hospitalisation in an ADR risk prediction model with an area under the receiver operator curve of 79.0% (95% CI 73.9%-84.1%). Most ADRs (106, 89.1%) were considered preventable. Another component of a prospective observational study determined the prevalence of mortality attributable to ADRs in patients presenting to hospital, and identified drugs and factors associated with ADR-related mortality. Of 1,001 patients, 15 (1.5%, 95% CI=0.80-2.30%) died. Deaths were primarily due to suspected drug-induced hepatotoxicity (7 patients, 43.8%) followed by acute kidney injury (4 patients, 25.0%). Anti-TB drugs and ART together were implicated in 60% of the deaths. A bivariate comparison showed patients who died with ADRs were more likely to have pre-existing liver disease (40.0% vs. 7.0%; 95% CI=8.1- 57.8%), a history of ADRs (40% vs. 1.4%; 95% CI=13.8-63.4%), a low BMI (17.6 ¬¨¬± 2.1 vs. 20.0 ¬¨¬± 2.9; 95% CI=0.9-3.9), exposure to anti-TB drugs (46.7% vs. 18.9%; 95% CI=2.3-53.1%) and ART (40.0 % vs. 7.7%; 95% CI=7.5-57.2%), a higher mean (¬¨¬±SD) number of medications (7.1 ¬¨¬± 3.3 vs. 3.8 ¬¨¬± 2.1; 95% CI=2.2-4.4), and Charlson Comorbidity Index (3.9 ¬¨¬± 2.9 vs. 1.6 ¬¨¬± 1.8; 95% CI=1.4-3.2) than surviving patients without ADRs. Findings from a series of analyses in a prospective observational study led us to further characterise the clinical patterns and severity of drug-induced hepatotoxicity (DIH), which was the commonest ADR implicated in hospitalisation and mortality. In this sub-study, 674 patients with documented previous medical history and regular medication prior to hospital admission and at least one set of liver function tests were included. Of 674 patients, 35 (5.2%) were deemed to have been hospitalised due to DIH, of whom, 22 (62.9%) exhibited a cholestatic pattern, 8 (22.9%) a hepatocellular pattern and 5 (14.3%) a mixed pattern. The most frequent drug classes implicated were anti-TB drugs (21 patients, 60.0%) followed by ART (12 patients, 34.3%). More than two-thirds of the DIH cases (24, 68.6%) were severe or fatal, were mainly caused by anti-TB drugs (15, 42.9%), ART (4, 11.4%) or concomitant anti-TB/ART (6, 17.1%). Our studies provided several novel findings regarding hospitalisation and mortality related to ADRs in Ethiopian patients. Our work revealed that the extent of ADR-related hospitalisation in adults is an important public health problem, with a significant number of fatal ADRs in patients presenting to hospital. Commonly used drugs, such as anti-TB drugs, ART and cardiovascular agents, causing well-known reactions are the most frequently occurring ADRs in patients presenting to hospital, suggesting that strategies for their prevention should be identifiable. Conversely, the ADR-related hospitalisation risk prediction model demonstrated some ability to identify patients at higher risk for ADRs, such as patients with lower BMI, previous ADR history, renal and liver diseases, multiple comorbidities and medications. This was further augmented by an ADR preventability assessment using Schumock and Thornton‚ÄövÑv¥s criteria, in which the majority of the ADRs were preventable provided these risk factors were reviewed and monitored closely. Therefore, consideration of the independent risk factors for ADRs identified in this study, by medical practitioners during assessment of patients at emergency and chronic care centres, might help distinguish patients who are at higher risk of ADR-related hospitalisation. More research is needed into intervention strategies to help reduce ADR-related hospitalisation and mortality. However, key areas that demand urgent interventions based on our study findings include patients taking anti-TB drugs (isoniazid and pyrazinamide) and ART (tenofovir, efavirenz and nevirapine), with a special focus on patients with malnutrition, previous ADR history, and pre-existing renal and liver diseases. Patients with cardiovascular disorders taking furosemide, enalapril, atorvas...