Original research
variation in resistance patterns and local stewardship priorities (UK-adapted version, see online supplemental material 1). 3 Although the WHO named AMR a public health priority for 2020, priorities changed with the emergence of the COVID-19 pandemic. Work conducted in England indicated a negative impact on an estimated 64% of stewardship activities when stewardship leads were surveyed. 4 Furthermore, it was hypothesised that COVID-19 may have increased prescribing of common respira- tory antibiotics given the similar presentation of COVID-19 and bacterial respiratory tract infections. 5 A meta-analysis found up to 70% of patients with COVID-19 received antibiotics, while the rate of bacterial coinfection was estimated to be 8.6%. 6 Subsequent studies have suggested that this figure is even lower, with a more recent study finding the rate of bacterial coinfection in patients with COVID-19 to be 3%. 7 These figures likely indicate high levels of inappropriate prescribing of antibiotics in patients with COVID-19 admitted to hospital early in the pandemic. In the UK, the National Action Plan for AMR (2019–2024) included targets for a 10% reduction in hospital use of Reserve and Watch antibiotics using 2017 as the baseline. 8 In Scotland, to meet WHO targets, at least 60% of overall antibiotic use should be Access group antibiotics by the end of 2022. 8 Nevertheless, there are currently knowledge gaps with regard to in-hospital prescribing of antibiotics. There is a lack of up-to-date data on utilisation patterns, especially in the era of COVID-19; moreover, data on the characteristics of hospitalised patients being prescribed anti- biotics are scarce. The Hospital Electronic Prescribing and Medi- cines Administration (HEPMA) system is being rolled out across Scotland and can provide inpatient, individual-level medicines data which may fill these data gaps. This study, therefore, aimed to evaluate the utilisation pattern of antibiotics according to the UK AWaRe classification using HEPMA data, including assessing any impact of the COVID-19 pandemic.
Statistical analysis All analyses were descriptive. Categorical variables were summarised as frequencies (count, percentage). For percentages, the numerator was the number of prescriptions of interest, and the denominator was the overall number of prescriptions of any antibiotic during the study period. Relative and absolute changes as well as the average quarterly change for each class were calcu- lated using trend analysis. Simple linear regression was used to determine significance of trends over time. Analyses have been conducted using R Studio Server Pro V.3.6.1. Reporting follows the Strengthening the Reporting of Observa- tional Studies in Epidemiology (STROBE) guidelines 9 ; a STROBE checklist can be found in online supplemental material 2. RESULTS The study comprised 1353003 unique prescriptions of antibi- otics across the 4-year study period. Of those, 702318 (51.9%) were for female patients; 582059 (43.0%) prescriptions were for 16–65year-olds; and 770944 (57.0%) were for >65 year-olds. The most prescribed antibiotics every year were amoxicillin (Access, 261562 prescriptions), followed by co-amoxiclav (Watch, 196902 prescriptions) and metronidazole (Access, 147033 prescriptions). When comparing year-on-year utilisation across AWaRe cate- gories, a steady absolute increase was apparent in proportions of Access antibiotics used, growing by 7% overall; from 55.3% (31 901/57 708) in Q1 2019 to 62.3% (106 449/170 995) in Q2 2022, a relative increase of 12.7%. A corresponding absolute decrease of 7.4% was observed in the watch class of antibiotics over the same time period from 42.9% (24 772/57 708) in Q1 2019 to 35.4% (60 632/170 995) in Q2 2022, a relative decrease of 17.3%. When stratified by sex, there were only minor differences. The aforementioned increases and decreases in Access and Watch classes, respectively, were observed equally in male and female patients, across all years studied (figure 1). No notable difference was seen in the proportion of Reserve antibiotics being used, continuing to represent a small percentage of overall use, ranging from a low of 1.4% in Q1 2020 to a high of 2.1% in Q1 2021. Use of Reserve antibiotics was higher every year in men than in women; however, the differences were small (ranging from 0.1% to 0.6% difference). No considerable change was seen in Q1 and Q2 of 2020 when COVID-19 initially caused increased hospital admissions and the first national lockdown. When inspected by age, similar trends were seen with Access usage increasing in 16–65year-olds as well as over 65 year-olds (figure 2). Differences were more pronounced in the older age group with a 9.4% increase over the 4 years, compared with only 3.7% in the 16–65year-olds, increasing from 56.4% (19 353/34 337) to 65.8% (64 387/97 815) and 53.7% (12 548/23 371) to 57.4% (44 185/76 956), respectively, between Q1 2019 and Q2 2022. The decrease in the proportion of Watch antibiotics used was more pronounced in the older age group. Utilisation of Watch antibiotics in >65 year-olds decreased year on year for corre- sponding quarters, with a total fall of 9.6% from a high of 41.9% (14 376/34 337) in Q1 2019 to 32.3% (31 568/97 815) in Q2 2022. A more modest decrease from 44.5% (10 396/23 371) to 39.8% (30 616/76 956) was noted in the 16–65 year-olds, with far more variation in this age group quarter to quarter. No discernible seasonal trend was noted in any class. The trend of increasing proportions of Access antibiotic use over time was significant only in the >65age group (p<0.05). In both male (p<0.05) and female patients (p<0.05) this trend was also significant. The decreasing trend in Watch antibiotics
PATIENTS AND METHODS Study design
This study was a repeated, cross-sectional study of all antibiotics prescribed to hospitalised patients from 1 January 2019 to 30 June 2022 using HEPMA. Start dates were chosen as 2019 presented a prepandemic baseline of antibiotic use and also coincided with data becoming available from some of the largest hospitals in the country providing significant increases in patient numbers. Scot- land has 14 territorial health boards, of which six contributed data: NHS Ayrshire & Arran; NHS Dumfries & Galloway; NHS Forth Valley; NHS Greater Glasgow & Clyde; NHS Lanarkshire; and NHS Lothian, collectively covering around 60% of the Scottish population. The roll-out of HEPMA has been gradual over time with differing timescales across health boards and hospitals, there- fore, data availability differs between sites. All data for this study were obtained from HEPMA, including dates (prescribed and administered), drug names and formulations, dosages and routes of administration. HEPMA also includes patient age and sex via unique patient identifiers. Data on antibiotics were extracted using the chemical name of the drug and route of administration. Study population and outcome The study subjects comprised prescriptions for any antibiotic indi- cated for systemic use and administered to patients >16 years, admitted to hospital for any reason during the study period. The study outcome was the antibiotic utilisation pattern in terms of quarterly number of prescriptions, classified according to the UK-adapted AWaRe classification, stratified by age groups and sex. 3
Proud E, et al . Eur J Hosp Pharm 2024; 0 :1–4. doi:10.1136/ejhpharm-2023-003874
2
4
BMJ1564-EAHP-Article-2.indd 4
26/02/2024 10:12
Powered by FlippingBook