JAC- Antimicrobial JAC Antimicrob Resist https://doi.org/10.1093/jacamr/dlad025 Resistance Antibiotic consumption trends in Ghana: analysis of six-years pharmacy issue data from a secondary healthcare facility Appiah-Korang Labi 1,2, Bridget S. Kartey3, George Kwesi Hedidor2, Benjamin Demah Nuertey2, Elsie Kodjoe2, Leslie No Vanderpuije4 and Noah Obeng-Nkrumah 5* 1Department of Medical Microbiology, University of Ghana Medical School, University of Ghana, Korle-Bu, Accra, Ghana; 2Life Course Cluster, WHO Ghana Country Office, No. 7 Ameda Street, Roman Ridge, Accra, Ghana; 3Pharmacy Department, Eastern Regional Hospital, Koforidua, Ghana; 4Pharmacy Unit, Ghana Health Service Eastern Regional Health Directorate, Koforidua, Ghana; 5Department of Medical Laboratory Science, School of Biomedical and Allied Health Sciences, University of Ghana, Accra, Ghana *Corresponding author. E-mail: nobengnkrumah@ug.edu.gh Received 27 September 2022; accepted 21 February 2023 Background: Surveillance of hospital antibiotic consumption provides data to inform corrective action and for monitoring antimicrobial stewardship activities. This study described antibiotic consumption patterns from 2016 through 2021 at a secondary healthcare facility in Ghana. Methods: Using the WHO methodology for surveillance of antimicrobial consumption in hospitals, we analysed a 6-year pharmacy issue data at the Eastern Regional Hospital. We report on the defined daily dose (DDD) per 100 patients, types of antibiotics consumed according to Anatomic Therapeutic Classification (ATC), WHO AWaRe classification; trends in antibiotic consumption and expenditure per DDD of antibiotics consumed. Results: Over the period, the mean (±standard deviation) antibiotic consumption rate was 256.7 ± 33 DDD/100 patients per year. A linear regression model showed an insignificant decreasing trend in antibiotic consumption (coefficient for time –0.561; P = 0.247). The top three consumed antibiotics in DDD/100 patients at ATC level 5 were amoxicillin-clavulanate (n = 372.6), cefuroxime (n = 287.4) and sulfamethoxazole-trimethoprim (n = 145.8). The yearly Access-to-Watch ratio decreased from 2.4 in 2016 to 1.2 in 2021. The mean yearly cost of antibiotics was $394 206 ± 57 618 US dollars. The top three antibiotics consumed in terms of cost were clindamycin $718 366.3, amoxicillin-clavulanate $650 928.3 and ceftriaxone $283 648.5. Conclusion: This study showed a sturdy rate of antibiotic consumption over the 6-year period with a year-on- year decrease in the Access-to-Watch antibiotic ratio. Data from pharmacy drug issues offer an opportunity to conduct antibiotic consumption surveillance at the hospital and national level in Ghana. Introduction well as optimize the use of antimicrobials.8 Optimizing antimicro- bial use and consumption through antimicrobial stewardship Globally, there has been a 65% increase in antimicrobial con- hinges on the availability of antimicrobial use and consumption sumption (AMC) in humans between the years 2000 and 2015, data. Antimicrobial use and consumption data can provide warn- with this increase driven mainly by low- and middle-income ing signs concerning antimicrobial exposure and use, allowing for countries.1 Increasing antimicrobial use is considered a major dri- the institution of corrective measures. It may also serve as a ver of antimicrobial resistance (AMR). AMR is considered a global monitoring tool for interventions aimed at reducing antimicrobial public health problem.2,3 exposure.1,9 In Ghana, surveillance of antimicrobial consumption Several reports from Ghana show increasing AMR, with a high at the hospital and national level is lacking, however, several prevalence of resistance phenotypes such as extended-spectrum point prevalence surveys on antimicrobial use have shown 47% beta-lactamases and increasing reports of carbapenemase- to 66% prevalence of antimicrobial use among hospitalized pa- producing bacteria.4–7 The Ghana national policy and action tients.10–13 Currently, Ghana is undergoing a national digitaliza- plan on AMR that is aligned with the global action plan on AMR tion agenda that includes the provision and use of electronic highlight the need to strengthen surveillance and research as health records in healthcare facilities. Electronic health records © The Author(s) 2023. Published by Oxford University Press on behalf of British Society for Antimicrobial Chemotherapy. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/ by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. 1 of 8 Downloaded from https://academic.oup.com/jacamr/article/5/2/dlad025/7081334 by University of Ghana. Balme Library user on 05 May 2023 Labi et al. of hospital pharmacies and doctors’ prescribing notes represent measuring drug consumption of different strengths or combinations and low-hanging fruit for conducting surveillance of antimicrobial can be used to compare rates between different antibiotic categories 19,20 use and consumption at the hospital level in Ghana. and years. The DDD was calculated by converting the total amount The Eastern Regional Hospital (ERH),14 is a secondary referral of antibiotic dispensed in into grams and divided by the standard WHO facility that provides a range of specialist services for the >2 mil- ATC DDD value given in grams. When measuring antibiotic consumption lion people living in rural and urban settings in the Eastern region in a hospital, where attendance data is available but no patient-level infor- 15 mation can be accessed, DDD per 100 patients is the recommended meth-of Ghana. The hospital has a history of electronic health records od for standardization.16 In this study, we used total patients’ attendance use for the past 6 years and has accumulated records useful for at ERH to represent the population served by the hospital. This was the conducting antimicrobial consumption studies. These data could closest and best representation of the population served by the hospital, provide an understanding of one of the potential drivers of AMR in as most in- and outpatients are accounted for in the patient attendance the hospital and inform antimicrobial stewardship strategies. In data. The DDD/100 patients were obtained by dividing the calculated total this study, we evaluated the antibiotic consumption trends at DDD for each year by the patient attendance for the respective year and the ERH from 2016 through 2021 using WHO’s defined daily multiplying the ratio by 100. The expenditure on antibiotic consumption dose (DDD) methodology. was calculated as expenditure per DDD by multiplying the number of anti- biotics issued in a period (e.g. all antibiotics or specific types) and the cost and dividing by the calculated DDD of the antibiotic for that period. For sub- Materials and methods group analysis, we compared yearly variations in antibiotic consumption for drug categories (e.g. AWaRe) by calculating the per cent contribution Study design and settings of that antibiotic category to the overall antibiotic consumption that year. Using the WHO methodology for the global programme on surveillance of antimicrobial consumption in hospitals,16,17 we analysed a 6-year phar- Data analysis macy central pharmacy issues data. The central pharmacy stores of the ® hospital dispense drugs to pharmacy units associated with various de- Data were entered and cleaned using Microsoft Excel 2021 and exported to ® partments of the hospital. The Pharmacy Department distributes drugs the STATA software for descriptive and analytical studies. Continuous vari- using the stock system, where dispensed drugs through the supply chain ables were presented as means ± standard deviations (SD) and sums. are recorded. Patients pay for any drugs consumed with cash or through Categorical variables were presented as frequencies and percentages. the National Health Insurance System. A recent point prevalence survey Difference in antibiotic consumption across the years was determined using conducted at the hospital showed an antimicrobial use prevalence of Chi-square test, with the Marascuilo’s post hoc procedure applied for subse- 50%.13 The hospital uses the national standard treatment guidelines as quent pairwise comparisons. The Chi-square trend analysis was used to the main document supporting antimicrobial prescriptions although compare the trend in consumption of WHO AWaRe antibiotic groups over other documents such as the British National Formulary are used.18 The the years. Measures of relative consumption, expressed as a percentage ERH has a microbiology laboratory that routinely performs bacterial cul- of the total consumption of groups of antibiotics, were derived for each anti- tures and antimicrobial susceptibility testing. The hospital’s Information biotic. A linear regression model was used to assess the trend of antibiotic Technology Department manages electronic health records where med- consumption rate over time. The coefficient for time and P value for the ical records of patients and pharmacy data are stored. trend of antibiotic consumption was calculated using yearly measures. All statistical tests were considered significant at a P value <0.05. The cumula- tive change in consumption for each antibiotic over the study period was cal- Antibiotic inclusion criteria culated by adding the absolute differences in DDD/100 patients between Data summaries on antibiotics dispensed over 71 consecutive months from 2017 and 2016, 2018 and 2017, 2019 and 2018, 2020 and 2019, and January 2016 through November 2021 were copied from pharmacy elec- 2021 and 2020. With regards to the cost of consumed antimicrobial agent tronic records and transferred to an Excel™-based data collection instru- in US dollars ($), the amount of antimicrobial agent in Ghana cedi was con- ment. Data extracted included the antibiotic name, the dosage, route of verted to dollars using the average yearly dollar rate. administration, quantity, date dispensed and the cost of the issued antibio- tics. No patient-level information on prescriptions—e.g. indication or age of Ethical considerations the patient—was available. Total patient attendance to the hospital for the years under consideration was retrieved. In this study, we included only Ethical clearance for the study was obtained from the Ghana Health data on antibiotics for systemic use, classified in the J01 category, by the Service Ethics Review Committee with protocol number GHS-ERC 004/ WHO Anatomic Therapeutic (ATC) classification system.17 The WHO also re- 05/22. All data extracted from the EHR were summaries and aggregated commends categorizing antibiotics as Access, Watch or Reserve data from antimicrobials issued from the pharmacy stores between 2016 (AWaRe),19,20 and we used the 2021 database20 to classify them accord- and 2021. No individual patient records were collected, thus findings from ingly. Topical antibiotics were excluded from the study. Antimicrobials the study cannot be linked to any patient. used for managing tuberculosis, parasites and fungi were not also included. Results Calculation of antibiotic consumption rates Over the 6-year survey period, we extracted data summaries on 15 Antibiotic consumption was estimated on the basis of the amount dis- different antibiotics at ATC level 4 (yearly mean ± SD, 14.33 ± 0.52) pensed from the central pharmacy store in standard units of milligrams and 26 antibiotics at ATC level 5 (yearly mean ± SD, 24.67 ± 0.82) or millilitres. Each standard unit was defined based on a single tablet, cap- sule, ampoule, vial or liquid preparation for oral consumption.16 Each anti- from the electronic records at ERH for analysis (Table 1). From a to- biotic was assigned to a WHO ATC level 4 and 5 classifications. We then tal of 26 types of antibiotics issued, 13 were in the Access category, expressed the number of drugs dispensed for each antibiotic class as 10 in the Watch category and none were in the Reserve category. our consumption data in DDD according to WHO methodology.16 The The antibiotics were issued to a yearly mean population of 170 DDD is the assumed average maintenance dose per day for a medicine 044.2 ± 10 878.5 patients attending the hospital. The mean vol- used for its main indication in an adult. It is a globally accepted unit for ume in DDD of antibiotics issued per year was 81 839.2 ± 190 2 of 8 Downloaded from https://academic.oup.com/jacamr/article/5/2/dlad025/7081334 by University of Ghana. Balme Library user on 05 May 2023 Antimicrobial consumption in secondary care hospital 3 of 8 Table 1. Summary of hospital metrics and daily defined doses 2016 2017 2018 2019 2020 2021 Summaries Hospital metrics Grand total Total patient visits (%) 155 560 (15.1) 161 084 (15.7) 177 969 (17.3) 192 428 (18.7) 163 336 (15.9) 177 639 (17.3) 1 028 016 (100) Hospital admissions (%) 16 350 (16.7) 16 348 (16.7) 16 340 (16.7) 16 326 (16.7) 16 324 (16.7) 16 332 (16.7) 98 020 (100) Number of ATC level 4 14 14 14 14 15 15 15 antibiotics prescribed Number of ATC level 5 24 25 24 26 24 25 26 antibiotics prescribed Distinct Access antibiotics 11 11 13 12 10 11 13 prescribed Distinct Watch antibiotics 8 9 10 8 8 9 10 prescribed Distinct Reserve antibiotics 0 0 0 0 0 0 0 prescribed Distinct unclassified 0 0 0 1 0 1 1 antimicrobials prescribed Volume in DDD (%) of antibiotics issued per year All antibiotics 397 276.6 497 341.3 514 471.3 422 362.0 361 231.0 437 103.4 2 629 786.0 Access antimicrobial 280 071.0 (70.5) 337 892.9 (67.9) 325 579.4 (62.4) 251 884.9 (59.6) 222 921.8 (61.7) 239 780.1 (54.9) 1 658 130.0 (63.3) Watch antimicrobials 117 205.6 (29.5) 159 448.4 (32.1) 188 891.9 (36.7) 170 469.6 (40.4) 138 309.2 (38.3) 197 283.2 (45.1) 971 608.0 (36.7) Reserve antimicrobials 0 0 0 0 0 0 0 Unclassified antimicrobials 0 0 0 7.5 (0.002) 0 40.0 (0.01) 47.5 (0.002) Oral antimicrobials 350 764.1 (88.3) 454 502.5 (91.4) 464 406.0 (90.3) 360 646.8 (85.4) 298 642.7 (82.7) 375 286.9 (85.9) 2 304 249.0 (87.7) Parenteral antimicrobials 46 512.5 (11.7) 42 838.8 (8.6) 50 065.3 (9.7) 61 715.2 (14.6) 62 588.3 (17.3) 61 816.4 (14.1) 325 536.4 (12.3) Generic antimicrobials 396 545.1 (99.8) 496 467.0 (99.8) 508 526.8 (98.8) 413 674.9 (97.9) 355 647.8 (98.5) 427 825.3 (97.9) 2 598 687 (98.9) Originator antimicrobials 731.5 (0.2) 874.3 (0.2) 5944.5 (1.2) 8687.1 (2.1) 5583.2 (1.5) 9278.0 (2.1) 31 098.6 (1.1) Paediatric formulation 25 797.5 26 633.8 28 363.3 24 635.2 13 427.3 25 552.2 144 409.4 (5.5) Adult formulation 371 479.1 470 707.5 486 108.0 397 726.7 347 803.7 411 551.2 2 485 376.0 (94.5) Access-to-Watch ratio 2.4 2.1 1.7 1.5 1.6 1.2 *%. Percentages: unclassified antimicrobials include Ceftriaxone sulbactam and Ciprofloxacin tinidazole. Downloaded from https://academic.oup.com/jacamr/article/5/2/dlad025/7081334 by University of Ghana. Balme Library user on 05 May 2023 Labi et al. (a) (b) (c) (d) Figure 1. Trends in antibiotic consumption rates in DDD/100 patient visits for: (a) all antibiotics, (b) oral versus parenteral antibiotics, (c) WHO AWaRe antibiotics and (d) Access-to-Watch antibiotic index with trend line that shows the WHO recommended ratio threshold of ≤1.6. 733.4. Adult antibiotics, compared to paediatric formulations, years. Overall, ∼94.5% (n = 1526.4/1543.9) of the total con- accounted for 94.5% (n = 1455.46/84.5) of the total volume in sumed antibiotics were generic brands. DDD of antibiotics issued over the 6 years. Antimicrobial consumption according to AWaRe Total antibiotic consumption rate classifications Figure 1(a) shows the antibiotic consumption in DDD/100 pa- Most (63.3%, n = 564.79/975.11) of the antibiotics consumption in tients for each year. The mean antimicrobial consumption DDD/100 patients over the 6 years belonged to the Access group rate for all antibiotics over the 6 years was 256.7 ± 33 DDD/ and there was no consumption of Reserve antibiotics over the period 100 patients. The highest consumption of 308.7 DDD/100 pa- of study. Unclassified antibiotics represented 0.026 DDD/100 patients tients was recorded in 2017 and the lowest of 219.5 DDD/100 accounting for 0.002% of total consumed antibiotics. Figure 1(c) patients was recorded in 2019. A linear regression model shows the trend in yearly antibiotic consumption in DDD/100 patients used to assess the pattern of antibiotic consumption over for the WHO AWaRe classification groups. There was a significant time showed an insignificant decreasing trend (coefficient for yearly increase (χ2 trend, P = 0.0001) in the consumption rate of time −0.561; P = 0.247). Differences in annual antibiotic con- Watch antibiotics from 2016 to 2021 and a commensurate progres- sumption levels were not significant for the 6 years under study sive decrease in consumption of Access class of antibiotics (χ2 trend, (χ2 = 6.409, P = 0.1706). When follow-up pairwise comparisons P = 0.0001). For example, in 2016, 70.5% (n = 180.1/255.4) of the to- were conducted with Marascuilo’s post hoc procedure, none tal DDD/100 patients belonged to Access group of antibiotics and this of the annual variations in antibiotic consumption between decreased to 54.9% (n = 135.0/246.1) in 2021. In contrast, the pro- any pair of years from 2016 to 2021 was significant (P > 0.05 portion of Watch antibiotics contributing to total DDD/100 patients for all 15 pairwise comparisons). Of the total antibiotics con- was 29.5% (n = 73.4/255.4) in 2016 and increased significantly sumed DDD/100 patients, 87.7% (n = 1354/1543.9) were oral over the years to 45.1% (n = 111.1/246.1) in 2021. The agents. Figure 1(b) shows a clear separation in which the con- Access-to-Watch ratio, based on antibiotic consumption rate in sumption trend in DDD/100 for oral antibiotics from 2016 DDD/100 patients, versus the year graph is displayed in Figure 1(d). through 2021 mirrors the consumption levels for all antibiotics There was a decrease in the Access-to-Watch ratio across the years [Figure 1(a)] and exceeds the consumption rate for parenteral from 2.4 in 2016 to 1.2 in 2021. Among the Watch group of antibio- antibiotics by several folds by a factor of >5 across the study tics, the top five with the highest consumption rate in DDD/100 4 of 8 Downloaded from https://academic.oup.com/jacamr/article/5/2/dlad025/7081334 by University of Ghana. Balme Library user on 05 May 2023 Antimicrobial consumption in secondary care hospital (a) (b) Figure 2. Antibiotic consumption at DDD/100 patients for Anatomic Therapeutic Classification (ATC): (a) Level 4 and (b) Level 5. *Comb. of pen incl BL, combination of penicillin including beta-lactams; Comb. of sulf. & trim., combination of sulphonanmides and trimethoprim derivatives; Sulf. & trim., sulphurmethoxazole and trimethoprim; Pen. with ext. spect, penicillin with extended-spectrum; BL, beta-lactamase; Ceft. BL combination, ceftriaxone beta-lacatamse combination; Phen. penicillin, Phenoxylmethyl penicillin; Amoxi.clav., amoxicillin-clavulanic acid. patients over the 6 years were cefuroxime (n = 287.4), azithromycin (n = 145.8), macrolides (n = 142.3) and imidazole derivatives (n = (n = 88.4), ciprofloxacin (n = 82.9), ceftriaxone (n = 41.9) and clarithro- 127.9). The complete list of antibiotics at ATC level 4 and their year- mycin (n = 39.9) (Table S1, available as Supplementary data at JAC ly consumption rates at can be found in Table S2. At ATC level 5, the Online). The top five Access antibiotics with the highest DDD/100 pa- top five antibiotics with the highest consumption rate in DD/100 tients were amoxicillin-clavulanate (n = 372.6), sulfamethoxazole- patients were amoxicillin-clavulanate (n = 372.6), cefuroxime (n = trimethoprim (n = 145.8), metronidazole (n = 127.8), doxycycline (n 287.4), sulfamethoxazole-trimethoprim (n = 145.8), metronidazole = 95.8) and clindamycin (n = 91.9). (n = 127.9), and doxycycline (n = 95.9) [Figure 2(b)]. Suffice to say that while amoxicillin-clavulanate remained the most con- Antimicrobial consumption rates by ATC classification sumed antimicrobial yearly, there were yearly variations in the positions of the other top five antibiotics as shown in Table S3. The DDD/100 patients according to ATC level 4 is presented in Figure 2(a). The top five consumed antibiotics are over the six-year period were penicillins with beta-lactamase combinations (n = Cumulative changes in antibiotic consumption 372.6.2), second-generation cephalosporins (n = 287.4), combina- Figure 3 shows the 6-year cumulative change in ATP level 5 anti- tions of sulfonamides and trimethoprim including derivatives biotic consumption. Cefixime and ceftazidime were the only 5 of 8 Downloaded from https://academic.oup.com/jacamr/article/5/2/dlad025/7081334 by University of Ghana. Balme Library user on 05 May 2023 Labi et al. Figure 3. Cumulative increase and decrease in antibiotic ATC level 5 consumption from 2016 to 2021. antibiotics with zero change in consumption DDD per 100 pa- the 6-year period were clindamycin ($718 366.3), amoxicillin- tients from 2016 through 20121. Other antibiotics including clavulanate ($650 928.3), ceftriaxone ($283 648.5), cefuroxime meropenem (−0.03 DDD/100 patients), vancomycin (0.001 ($262 708.9) and metronidazole ($80 258.5). DDD/100 patients) and amikacin (0.16 DD/100 patients) dis- played only marginal changes in DDD/100 patients. Azithromycin showed the highest net increase in consumption Discussion (26.78 DDD/100 patients), followed by amoxicillin-clavulanate Globally, antibiotic consumption has increased over the last two (11.84 DDD/100 patients), cefuroxime (8.92 DDD/100 patients) decades, fuelled mainly by consumption in LMICs.1 Data on anti- and ceftriaxone (7.60 DDD/100 patients). In contrast, other anti- biotic consumption are critical in informing and evaluating anti- biotics recorded a net decrease in consumption. The top four anti- microbial stewardship programmes implemented at the biotics with the highest drop in consumption were hospital and national levels to combat AMR. This study found sulfamethoxazole-trimethoprim (−47.33 DDD/100 patients), an insignificant drop in antibiotic consumption from 2019 to erythromycin (−7.65 DDD/100 patients), metronidazole (−6.42 2021 compared to 2016 to 2018, with the most consumed anti- DDD/100 patients) and benzyl-penicillin (−3.17 DDD/100 microbial in the hospital being amoxicillin-clavulanate. patients). Antibiotic consumption in the hospital increased from 2016 to 2018 and decreased sharply between 2019 and 2020 before be- Cost of antimicrobials consumed ginning to rise again in 2021. Whereas the drop in 2020 antibiotic consumption may be attributed to the COVID-19 pandemic The total cost of antibiotics over the 6 years was $2 365 241 with and the consequent drop in hospital attendance, it is not easy a yearly mean cost of $394 206 ± 57 618. The least gross amount to fathom the sharp drop in antibiotic consumption in 2019. of $301 011.7 was spent on antibiotics in 2020 and the highest One possible reason could be stockouts; however, this study did amount of $494 934 was spent in 2018. The yearly average not have access to this data. Amoxicillin-clavulanate and cefur- cost of antibiotics per DDD was $0.89 (SD 0.06), ranging from oxime were the two commonest antibiotics consumed in the $0.83 in 2020 to $0.95 in 2016. Patients attending the ERH spent hospital over the study period. Similarly, these agents have a mean of $2.3 ± 0.3 on antibiotics per visit to the hospital. The been documented as part of the top five antibiotics used in hos- top five antibiotics contributing to overall antibiotic cost over pitalized patients from previous point prevalence surveys in 6 of 8 Downloaded from https://academic.oup.com/jacamr/article/5/2/dlad025/7081334 by University of Ghana. Balme Library user on 05 May 2023 Antimicrobial consumption in secondary care hospital Ghana.10,13 These antibiotics are recommended for the treat- This could explain the lack of data on reserve antibiotics that were ment of upper and lower respiratory tract and urinary tract infec- not stocked by the hospital. Also, the data collected did not ac- tions by the Standard Treatment Guidelines of Ghana,18 and they count for stockouts that could have affected consumption levels. are on the national health insurance of Ghana’s list of drugs,21 so However, findings from this study are useful because it is one of are not prohibited by out-of-pocket cost. On the contrary, agents the first to describe hospital-wide antibiotic consumption in like ceftriaxone which feature prominently as part of the top five Ghana and describes the potential use of drug issuance data as antibiotics used in most point prevalence surveys in Ghana was a measure of hospital-level antibiotic consumption. These data the tenth most consumed antibiotic in this study.10,13 Also, there could serve as baseline data for future studies and antibiotic was sharp rise in the consumption of azithromycin in 2021. This stewardship initiatives. could be due to increased use as a result of speculations that sug- This study has the following policy implications. First, data gested azithromycin as an agent for treating COVID-19. This high- from pharmacy drug issues may be a useful resource for lights the important role of antibiotic consumption surveillance or conducting antibiotic consumption surveillance at the hospital studies in its ability to give a complete antibiotic utilization picture and national level and may represent low-hanging fruit for con- compared to point prevalence surveys. It is of interest to note, sumption surveillance in Ghana considering the EHR rollout however, that other antibiotics such as clindamycin and ceftriax- across healthcare facilities. Second, there is a need to understand one were responsible for the first and third antibiotics when ex- the reasons behind the decreasing Access-to-Watch ratio ob- penditure was assessed. served in this study through the conduct of further studies. This Overall, most antibiotics prescribed belonged to the Access will be important in influencing deliberate policies aimed at re- group of antibiotics, however, the proportion of Watch antibiotics versing the trend towards >60% prescription of antibiotics in consumed yearly increased from 2016 to 2021 and is exemplified the hospital belonging to the Access group.20 Third, there is by a decreasing Access-to-Watch index from 2.4 in 2016 to 1.2 in also the need to strengthen AMR surveillance activities at the 2021. This means that as of 2021 antibiotic consumption in the hospital to monitor the impact of the increase in the use of hospital did not conform to the WHO target, which requires at Watch antibiotics. least 60% of overall antibiotic consumption to be made of the Access group of antibiotics.20 This rise in Watch antibiotic use relative to the Access group of antibiotics is similar to global find- Conclusion ings especially those from lower- and middle-income countries This study showed a sturdy rate of antibiotic consumption over (LMICs) and may be a reflection of increasing antibiotic resistance the 6-year study period with a year-on-year decrease in the observed in the study setting as has been suggested elsewhere.22 Access-to-Watch ratio of antibiotics. Amoxicillin-clavulanate, Increasing the use of Watch antibiotics could also be attributed cefuroxime and sulfamethoxazole-trimethoprim were the three to improved economic status in many LMICs with an associated most consumed antibiotics; however, clindamycin, amoxicillin- increased purchasing power for more expensive broad-spectrum clavulanate and ceftriaxone were the top three antibiotics antibiotics,1 and uncertainty concerning febrile illness diagnosis. according to the total budget spent. Data from pharmacy drug 23 There was no consumption of reserve antibiotics in the hospital issues offer an opportunity to conduct antibiotic surveillance at over the study period, this may be a result of their unavailability the hospital level in Ghana and when aggregated may give a na- on the local market, absence in the Standard Treatment tional picture. This could be a useful resource for monitoring anti- Guidelines of Ghana18 and the fact that they are not funded by biotic stewardship activities at the hospital and national levels. the health insurance scheme 21. Similar absent use of reserve antibiotics has been documented in point prevalence surveys among hospitalized patients in Ghana.10,13,24,25 Acknowledgements This study has potential limitations. First, the study is limited to We thank Mr Ibrahim Fuseini of the IT department of the Eastern Regional one secondary healthcare facility thus finding may not be gener- Hospital for extracting data from the EHR. alizable to other healthcare settings. Second, the data could not be segregated into outpatient and inpatient status that would have been useful to show any differences in antibiotic consump- tion patterns. However, our data showed that 90% of antibiotics Funding were oral formulations, which suggests that a large proportion of This study did not receive any funding and was carried out as part of our consumed antibiotics was prescribed to outpatients. Antibiotic routine work. consumption data for 2021 was short by 1 month due to a change in EHR at the hospital, and this may have affected the overall antibiotic consumption observed especially for that year. Transparency declarations The level of antibiotic resistance in the population under study The authors declare no conflict of interests. is unknown and this could have given better meaning to the ob- served antibiotic consumption. Also, the findings in this study are Author contributions not commensurate with the appropriateness of use. This study A.K.L., B.S.K., G.K.H. and N.O.N. were responsible for concept and protocol used hospital antibiotics issue data that are dependent on the development; A.K.L. and B.S.K. obtained the data; A.K.L., N.O.N. and B.D.N. medications stocked by the hospital and did not account for performed data analysis; A.K.L. wrote the first draft and R.Z., L.N.O.V., E.K., medications that were prescribed and purchased in pharmacies G.K.H. reviewed the manuscript. All authors have read and agreed to the outside the hospital for use in the hospital by individual patients. final version of the manuscript. 7 of 8 Downloaded from https://academic.oup.com/jacamr/article/5/2/dlad025/7081334 by University of Ghana. Balme Library user on 05 May 2023 Labi et al. 11 Labi A-K, Obeng-Nkrumah N, Sunkwa-Mills G et al. Antibiotic prescrib- Availability of data ing in paediatric inpatients in Ghana: a multi-centre point prevalence The data for this study are available upon reasonable request from the survey. BMC Pediatr 2018; 18: 391. https://doi.org/10.1186/s12887-018- corresponding author. 1367-5 12 Labi A-K, Obeng-Nkrumah N, Owusu E, et al. Multi-centre point prevalence survey of hospital-acquired infections in Ghana. J Hosp Supplementary data Infect 2019; 101: 60–68. https://www.sciencedirect.com/science/article/ pii/S0195670118302573. https://doi.org/10.1016/j.jhin.2018.04.019 Tables S1 to S3 are available as Supplementary data at JAC Online. 13 Labi A-K, Obeng-Nkrumah N, Dayie NTKD et al. Antimicrobial use in hospitalized patients: a multicentre point prevalence survey across seven hospitals in Ghana. JAC-Antimicrob Resist 2021; 3: dlab087. https://doi. References org/10.1093/jacamr/dlab087 1 Klein EY, Van Boeckel TP, Martinez EM et al. Global increase and geo- 14 Eastern Regional Hospital. https://erhk.org/index.html. graphic convergence in antibiotic consumption between 2000 and 15 Ghana Statistical Services. https://www.statsghana.gov.gh/regional 2015. Proc Natl Acad Sci U S A 2018; 115: E3463-70. https://doi.org/10. population.php?population=MTM5ODc0NTI3OS45NTQ1&&Eastern& 1073/pnas.1717295115 regid=5, 2023. 2 Holmes AH, Moore LS, Sundsfjord A et al. Understanding the mechan- 16 World Health Organization. WHO methodology for a global pro- isms and drivers of antimicrobial resistance. Lancet 2016; 387: 176–87. gramme on surveillance of antimicrobial consumption. https://apps. https://doi.org/10.1016/S0140-6736(15)00473-0 who.int/iris/bitstream/handle/10665/336215/9789240012639-eng.pdf, 3 Goossens H. Antibiotic consumption and link to resistance. Clin 2020 Microbiol Infect 2009; 15: 12–5. https://doi.org/10.1111/j.1469-0691. 17 WHO. Guidelines for ATC Classification and DDD Assignment. WHO 2009.02725.x Collaborating Centre for Drug Statistics Methodology, 2013. 4 Banu RA, Alvarez JM, Reid AJ et al. Extended spectrum beta-lactamase 18 Ministry of Health Ghana National Drugs Programme. Standard Escherichia coli in river waters collected from two cities in Ghana, 2018– Treatment Guidelines. Seventh Edn. Accra, Ghana, 2017. 2020. Trop Med Infect Dis 2021; 6: 105. https://doi.org/10.3390/ 19 Hsia Y, Lee BR, Versporten A et al. Use of the WHO access, watch, and tropicalmed6020105 reserve classification to define patterns of hospital antibiotic use 5 Codjoe FS, Donkor ES, Smith TJ et al. Phenotypic and genotypic charac- (AWaRe): an analysis of paediatric survey data from 56 countries. terization of carbapenem-resistant gram-negative Bacilli pathogens from Lancet Glob Health 2019; 7: e861–71. https://doi.org/10.1016/S2214- hospitals in Ghana. Microb Drug Resist Larchmt N 2019; 25: 1449–57. 109X(19)30071-3 https://doi.org/10.1089/mdr.2018.0278 20 WHO. 2021 AWaRe classification. https://www.who.int/publications- 6 Labi A-K, Nielsen KL, Marvig RL et al. Oxacillinase-181 carbapenemase- detail-redirect/2021-aware-classification, 2021. producing Klebsiella pneumoniae in neonatal intensive care unit, Ghana, 21 Medicines List. http://www.nhis.gov.gh/medlist.aspx, 2023. 2017–2019. Emerg Infect Dis 2020; 26: 2235–8. https://doi.org/10.3201/ 22 Klein EY, Milkowska-Shibata M, Tseng KK et al. Assessment of WHO eid2609.200562 antibiotic consumption and access targets in 76 countries, 2000–15: an 7 Obeng-Nkrumah N, Twum-Danso K, Krogfelt KA et al. High levels of analysis of pharmaceutical sales data. Lancet Infect Dis 2021; 21: extended-spectrum beta-lactamases in a major teaching hospital in 107–15. https://doi.org/10.1016/S1473-3099(20)30332-7 Ghana: the need for regular monitoring and evaluation of antibiotic resist- 23 Adrizain R, Setiabudi D, Chairulfatah A. The inappropriate use of anti- ance. Am J Trop Med Hyg 2013; 89: 960–4. https://doi.org/10.4269/ajtmh. biotics in hospitalized dengue virus-infected children with presumed con- 12-0642 current bacterial infection in teaching and private hospitals in Bandung, 8 Ministry of Health. Ghana National Action Plan on antimicrobial resist- Indonesia. PLoS Negl Trop Dis 2019; 13: e0007438. https://doi.org/10. ance (2017–2021). 2017. https://www.moh.gov.gh/wp-content/uploads/ 1371/journal.pntd.0007438 2018/04/NAP_FINAL_PDF_A4_19.03.2018-SIGNED-1.pdf 24 D’Arcy N, Ashiru-Oredope D, Olaoye O et al. Antibiotic prescribing pat- 9 Saleem Z, Hassali MA, Godman B et al. Point prevalence surveys of anti- terns in Ghana, Uganda, Zambia and Tanzania hospitals: results from the microbial use: a systematic review and the implications. Expert Rev Anti global point prevalence survey (G-PPS) on antimicrobial use and steward- Infect Ther 2020; 18: 897–910. https://doi.org/10.1080/14787210.2020. ship interventions implemented. Antibiotics 2021; 10: 1122. https://doi. 1767593 org/10.3390/antibiotics10091122 10 Labi A-K, Obeng-Nkrumah N, Nartey ET et al. Antibiotic use in a ter- 25 Amponsah OKO, Buabeng KO, Owusu-Ofori A et al. Point prevalence tiary healthcare facility in Ghana: a point prevalence survey. Antimicrob survey of antibiotic consumption across three hospitals in Ghana. Resist Infect Control 2018; 7: 15. https://doi.org/10.1186/s13756-018- JAC-Antimicrob Resist 2021; 3: dlab008. https://doi.org/10.1093/jacamr/ 0299-z dlab008 8 of 8 Downloaded from https://academic.oup.com/jacamr/article/5/2/dlad025/7081334 by University of Ghana. Balme Library user on 05 May 2023