Infection prevention and control has played key roles in the fight against the COVID-19 pandemic, as indicated by the absence of any infections among the nearly 42,000 medical support staff in Hubei Province. The simultaneous increase in the public's awareness of infection prevention and control measures was also successful in preventing COVID-19 in the community . Although China's domestic policies were successful, the pandemic is still a serious problem in many other countries, and China still has a risk that new cases will be imported. The “Normalized Epidemic Prevention and Control Requirements” measures of May 2020 have been key to the control of the COVID-19 pandemic [7, 8]. Based on risk classification, targeted measures should be taken, such as wearing masks in densely closed places. To prevent routine infection and improve infection control, hospitals have taken additional epidemic prevention and control measures, especially wearing masks and strengthening the regulations regarding individuals entering and exiting hospitals.
It is now more than 1 year after implementation of the “Normalized Epidemic Prevention and Control Requirements”, and this study showed that the HAI rate in 2020 (1.82%) was similar to the average HAI rate reported during the most recent five years in China (1.94%) , but significantly higher than the HAI rates at our facility during 2018 and 2019. Although the CAI rates in 2020 and 2019 were significantly higher than in 2018, there was no difference between 2019 and 2020. Our results indicated the lower respiratory tract was the most common site of HAIs, were similar to the results previously reported for Beijing  and in a 5-year national cross-sectional survey . This may be because the respiratory tract is more vulnerable to HAIs, and because it is easy to access the respiratory tract for collection of specimens and diagnosis. We found no significant differences in the percentage of lower respiratory tract HAIs from 2018 to 2020.
From the perspective of epidemic prevention and control, measures such as wearing masks and maintaining a proper social distance are important for prevention of respiratory infectious diseases. However, the rates of HAIs of the respiratory system did not change during our 3-year study period. In addition, measures such as fixed escort, restricted visits, and public education [18–19] can reduce the risk of cross-infection, but our results provided no evidence that they prevented HAIs. There is a general consensus that increased hand hygiene compliance can reduce the incidence of HAIs [20, 21] and that overall use of alcohol-based hand cleaner is a reliable indicator of hand hygiene compliance [22, 23]. We found that the use of hand cleaner and average number of open beds per day increased in tandem from 2018 to 2020, but this apparently had no impact on the rate of HAIs.
Analysis of the causes and occurrence of HAIs can be difficult  because of the impact of endogenous and exogenous factors. Thus, even if exogenous factors are controlled through prevention and control measures, the effects of endogenous factors may remain. Moreover, as a regional comprehensive medical center, our institution mostly admits emergency and critical patients, and a patient's condition determines the risk of a HAI, but the prevention and control measures at the hospital do not.
Our analysis of the anatomical sites of CAIs indicated the lower respiratory tract was the most common site (as with HAIs), but the percentage of lower respiratory tract infections was lower in 2020 than during the previous two years. This may be related to the implementation of the “Normalized Epidemic Prevention and Control Requirements” at the community level. In particular, the COVID-19 pandemic forced people to avoid unnecessary socialization, to wear masks, and to perform frequent hand washing to reduce the probability of cross-infection. However, our analysis of upper respiratory tract CAIs indicated no significant change over time. This may be because most upper respiratory tract infections are caused by viruses, and these patients often recover without the need for hospitalization. Therefore, our percentage of CAIs of the upper respiratory tract among inpatients (less than 2%) was probably much lower than the percentage in all community-dwelling individuals.
The composition of pathogens responsible for HAIs was relatively stable from 2018 to 2020. Gram-negative bacteria accounted for nearly 60% of the top ten species, and K. pneumoniae ranked first; in contrast, other studies in China reported that P. aeruginosa ranked first among HAIs [15, 17, 25]. This difference may be due to geographic differences. The pathogens responsible for CAIs were similar to those responsible for HAIs (i.e., mainly Gram-negative bacteria), but E. coli ranked first for CAIs. This is probably because the urinary system was a much more common infection site for CAIs. It is particularly noteworthy that proportion of fungal CAIs and HAIs increased from 2018 to 2020. This may be due to the increase of opportunistic infections caused by the increasing incidence of tumor diseases and use of immunosuppressants . Our analysis of MDROs indicated no significant changes, except that MRSA infection was significantly greater during 2020 than 2018 and 2019. This may be because HAIs by MDROs mostly occur in the ICU , and most of these patients have acute and critical diseases, poor clinical status, and the measures of the “Normalized Epidemic Prevention and Control Requirements” had no impact in preventing infections by MDROs in this specific population.
The rational use of antibiotics is closely related to patient prognosis and the occurrence by antibiotic-resistant bacteria. The problem of drug-resistant bacteria has become a serious worldwide public health problem. The frequency of antibiotic use, the culture rate of microbial specimens, duration of antibiotic use, and AUD are all important indicators . Our results showed that from 2018 to 2020, the rate and duration of use gradually decreased, and the rate of culturing before use had a gradual increase. Although the utilization rate decreased, administration of antibiotics to more than 20% of all inpatients was still very high because the sum of all HAIs and CAIs among inpatients was less than 5%. At the same time, the significant declines in the utilization rate and days of utilization may be related to changes in Chinese medical insurance policies, in addition to the increased awareness by doctors of the problem of over-prescribing antibiotics. AUD is a key index used to assess the rationality of the long-term use of antibiotics. Previous studies confirmed that AUD is closely related to bacterial drug resistance [29, 30]. The present study showed that AUD increased slightly (but not significantly) from 2018 to 2020, consistent with the slight increases in HAIs and CAIs. Our analysis of the AUD of individual drugs indicated the AUDs of the five most common drugs (mostly cephalosporins) were relatively stable from 2018 to 2020, similar to the results of Yao et al. . This may be because of the greater convenience and safety of these drugs, and the fewer restrictions regarding their use . In addition, changes in the AUD of antibiotics may also be related to changes in infection sites, such as increasing proportion of infections by Gram-negative bacteria and fungi.
Of course, there are some limitations in this study. First, the study period was only 3 years, and may not be fully representative. Second, this study was conducted at a single center study, and the research area and scope need to be expanded to verify the generalizability of the conclusions.