In this study, we estimate secular trends of antibiotic utilisation in China’s secondary and tertiary hospitals over an 8-year period, using procurement data from a national sampling database and the accepted ATC/DDD methodology. This allows us to analyse the antibiotic utilisation in China’s hospital setting and benchmark it internationally, which would be informing for healthcare providers, decision-makers, as well as the public.
In the nearly past two decades, the Chinese government has attempted to confine AMR by a series of policies and measures including limiting antibiotic use. The evolution of the policy management has been well documented [10]. According to the National Health Commission, the outpatient antibiotic prescription proportion in Chinese hospitals (secondary and tertiary hospitals) has continuously declined from 16.2% in 2011 to 7.7% in 2017. Similarly, the inpatient antibiotic prescription proportion has also declined from 55.2% to 38.0% [18]. In addition, studies were conducted on the impact of antibiotic use in secondary and tertiary hospitals, analysing the effect of various drug management programs on promoting the proper use of antibiotics. It should be noted that the implementation of the zero mark-up policy cut off the economic incentives between the volume of prescriptions and the hospital revenue [13, 19, 20]. However, our study showed that antibiotic utilisation in China’s hospitals increased significantly (38.2% increase between 2011 and 2018). Although the developed regions with larger populations consumed more antibiotics, significant increases were observed in underdeveloped regions despite having smaller populations (4 out of the top 5 provinces with the highest increase in antibiotic utilisation were underdeveloped). While it can be hard to explain the discrepancy with other studies that describe a decrease in antibiotic use [21-24], we found evidence that is consistent with the increasing tendency, when considering changes in the pharmaceutical market and inpatient volume. China’s anti-infection pharmaceutical market continuously increased from 2011 to 2017 with varying growth rates. Meanwhile, the inpatient volume at China’s tertiary hospitals nearly tripled between 2010 to 2017 according to China Health Statistics Yearbook [17]. This indicated that the demand for antibiotic treatment was growing. What’s more, although the proportion of antibiotic prescriptions was decreasing, the median antibiotic usage intensity in 177 core members of the Centre for Antibacterial Surveillance was slightly increasing from 2014 to 2016 from 47.21 to 47.65 DDDs/100 patient days. In addition, the methodology in our study was different from the one used in the national report. The difference stems from the latter report extracting data from the Centre for Antibacterial Surveillance, which is a national surveillance network for collecting antibiotic prescription data across more than 1300 hospitals nationwide. This might indicate that the composition of the sample hospitals between CMEI database and national surveillance network was different. However, due to the lack of access to the list of sample hospitals, a broader study with more hospitals covered is needed to verify the difference.
When compared with European countries, our results showed that antibiotic utilisation in China’s hospitals was more than three times than the average level (mean) of antibiotic consumption in the hospital setting in 24 EU/EEA countries (2.0 DID in 2017, ranged from 0.9 to 3.1 DID) [25]. One more worrying fact was that the antibiotic utlisation in China’s hospitals increased by a CAGR of 4.3% while the same metric, taken on an average annual basis in EU/EEA countries’ hospital settings was 1.0% [25]. However, these results need to be interpreted with caution, because in EU/EEA countries outpatient antibiotic use accounted for 90%–94% of the total (considering countries that provided separate data in ESAC) [25, 26], whilst the majority of antibiotics were consumed in a hospital setting in China. This might explain why the antibiotic utilisation in China’s hospitals was lower than that of antibiotics consumed in the community setting in EU/EEA countries (6.3 DID vs 21.8 DID in 2017) [25]. AWaRe categories are proposed by WHO in the context of a comprehensive review of the optimal antibiotic choices for many common infectious syndromes in adults and children. WHO set up a global target of having the proportion of Access antibiotics be greater than 60% of all antibiotics prescribed to reduce AMR [27]. Our study showed that the Access category proportion in the sample hospitals was only 19.4%, which was lower than the proportion in most countries that collected this data [27]. The massive use of second- and third-generation cephalosporins, macrolides, as well as combinations of penicillins contributed to the high proportion of the Watch category. Measuring antibiotic utilisation by quantifying the antibiotic use in each of the AWaRe categories allows some inference about the overall quality of antibiotic use between countries. The combination of both absolute and relative utilisation by category allows simple benchmarking (e.g. an overuse of Watch antibiotics can become immediately apparent and a reduction in Watch antibiotics can be identified as a target for antibiotic stewardship interventions) and assessment of trends over time (to evaluate the impact of interventions) [2].
When looking at the pattern of antibiotic utilisation, we found that cephalosporins were continuously the most consumed antibiotics in China’s hospitals, followed by combinations of penicillins, which were often defined as extended-spectrum antibiotics, and quinolones. This pattern was similar to the previous studies conducted in other regions of China [19, 20], as well as national surveillance data [18]. Unlike in China, more penicillins were prescribed in Europe and the United States [28-30]. This might partly be attributed to the different settings that antibiotics were consumed: mostly inpatient in China compared to an ambulatory setting in Europe and the US. Cephalosporins were recommended by the national guidance for the majority of the perioperative prophylaxis settings in China [31]. Although quinolones were recommended by US FDA to those who have no other treatment options [32], Chinese physicians may prefer quinolones and cephalosporins more than penicillins due to the time-consuming skin allergy testing requirements for penicillins prior to administration [33]. Although the utilisation of carbapenems, a class of last-resort antibiotics, were similar to EU/EEA countries (0.10 DID vs 0.06 DID in 2017, country range: 0.02-0.17), the significant increase of carbapenems utilisation still cannot be ignored (0.03 DID in 2011 to 0.10 DID in 2018) [25]. Because carbapenems are categorized as highly restricted antibiotics by the National Health Commission and require pre-authorization before use in China, this increase could be partly due to the rise in extended-spectrum β-lactamase-producing Gram-negative bacteria, which has been identified in epidemiological surveillance studies [18, 34]. Alongside the increase of carbapenem utilisation, the resistance rate of carbapenem-resistant Enterobacteriaceae has also been reported [35].
Population-weighted antibiotic utilisation was greater in secondary hospitals than in tertiary hospitals. As more severe patients were admitted in tertiary hospitals compared with secondary hospitals, we speculated that more potential inappropriate use of antibiotics might exist in lower level hospitals. Studies found that diseases, such as diarrheal illness, colds, pharyngitis, acute bronchitis, were likely to be prescribed antibiotics in rural and underdeveloped regions [36-38], although most of these illnesses are viral instead of bacterial. This is consistent with our results that most of the increases in antibiotic utilisation were observed in underdeveloped regions. However, further study on the level of individual antibiotic usage, such as a manual medical record review, is needed to confirm our speculation.
The findings of this study are subject to several limitations. First, the hospitals in the database were on a voluntary basis instead of mandatory participation, especially the proportion of secondary hospitals in the study was relatively low, therefore could bring selection bias. Besides, the sample hospitals in our study were all public hospitals and the sample size was relatively small, however, since the number of patient visits in public hospitals was more than 85% of total patient visits in China [17], and the CMEI database covered about 40% of total drug procurement at city-level public hospitals, therefore it still made our study to be representative to a certain extent. Second, the population denominator used in the study was determined under certain conditions which may underestimate antibiotic utilisation as it cannot include cross-provincial patient flow. Third, since we cannot access bed day data and the size of the hospital, the estimation of the population covered by the sample hospitals might be biased. Finally, as the study analysed procurement data rather than clinical usage of antibiotics, we were unable to determine the appropriateness of antibiotic use at the individual level.