Combined Use of Antibiotics as a Risk Factor for Health Care–Associated Infections: A Case-Control Study

are the most frequently prescribed medications. Many studies suggest an increased risk of health care–associated infections (HAIs) among However most related have We conducted a determine the association

from infectious diseases specialists, the Centers for Disease Control and Prevention, the World Health Organization, and U.S. and European governments [10][11][12]. The major risk factors associated with HAIs were low albumin, tracheostomy, prior hospital stay, central venous catheter insertion, urinary system disease, high blood glucose are associated, and intensive care unit (ICU) admission [13][14][15]. In addition, antibiotic exposure has been considered a risk factor for HAIs in previous studies [11,16,17]. However most related studies have focused only on a single type of HAIs, but given the complexity and extent of the subject, it is necessary to conduct extensive research. Until recently, there is some lack of knowledge about the relationship between combined antibiotic exposure and HAIs. A case-control study was undertaken to answer this question.
The aim of this study was to comprehensively analyze the association between antibiotics or combination of antibiotics and the risk of HAIs in a retrospective case-control study.

Setting
This study was carried out at a liated hospital of Northwest Minzu University, a 1010-bed university teaching hospital in Northwest China. All 216 patients identi ed as HAIs between 1st January 2019 and 31st December 2019 (cases) were identi ed from report cases. Sixteen patients were excluded which did not meet the diagnostic criteria for HAIs.
Controls were selected at random from the patients who were not diagnosed with HAIs (the controls are strati ed according to the quarterly distribution of cases and the sampling rates for the rst quarter, second quarter, third quarter, and fourth quarter are 25%, 15%, 20%, and 40%, respectively) and were matched to cases based on department( at a ratio of 1:2)( Figure 1). Controls were excluded if it meets the diagnostic criteria for HAIs when viewing the cases. Case patients and controls were also excluded if their case data were incomplete. Total of 428 quali ed controls were selected. Clinical datas were extracted from paper and electronic hospital case notes, and inspection records. The study was approved by the a liated hospital of Northwest Minzu University Research Ethics Committee.

Exposures
Documented prescription of antibiotics (Contains β-lactams, macrolides, lincomycin, peptides and quinolones, etc.) or antibiotic combination (the simultaneous or sequential use of two or more antibiotics) on admission and before of hospital infection diagnosis was recorded. Antibiotics taken 1 to 2 days before the diagnosis of HAIs were not analyzed because it is in the incubation period of potential infection.

Data collection
Datas were obtained from patients' medical records, and relative datas were recorded on structured abstraction forms. Variables analyzed as possible confound included demographics (age, sex, marital status, ethnic, education and hospital length of stay); invasive procedures (urinary catheter insertion, mechanical ventilation, etc.) during hospitalization; exposure (greater than two day) to antimicrobials (βlactams, macrolides, lincomycin, peptides and quinolones, etc.) after admission and before the occurrence of nosocomial infection.

Statistical analysis
Demographic characteristics of the cases and controls were presented using descriptive statistics.
Continuous variables were presented as mean ± SD, and we used t-tests for comparisons. As the results of the length of hospital stay of the datas for the two groups showed non-normal distribution, they were compared with the median, and the data for two groups were compared using the Wilcoxon rank-sum test. We presented categorical variables as numbers and percentages, and compared percentages using the chi-square test. Multivariate logistic regression models were used to compare each case group and control group. A forward elimination process was used, and adjusted odds ratios and 95% con dence intervals were calculated. Results of the regression models are presented as ORs along with 95% con dence intervals (CIs). A factor was considered to be statistically signi cant if the 95% CI of the corresponding OR did not contain 1.
A two-tailed P value of less than 0.05 was considered to show statistical signi cance, and statistical analyses were performed using SPSS Statistics 23 (IBM Corp, Armonk, NY, USA).

Patient involvement
No patients were involved in the design or implementation of the study, or writing up the results.

Results
During the study period, 216 cases were identi ed and selected with 428 controls for analyses. The clinical characteristics of these patients are shown in Table1. The mean age of case patients (66 years) was higher than that of the control groups (60years). The gender, ethnic, education, and marital status were similar for case patients and controls. The median hospital length of stay of case patients(20days) was longer than that of control groups(10days).The percentages of patients with a urinary catheter ,Ventilator or central catheter (urinary catheter, 42.6%; Ventilator,19.9%;central catheter, 13.9%) were higher in the cases than in the control groups (urinary catheter,7.5%;Ventilator, 0.7%;central catheter,0.7%) (P<0.001 for the three comparisons).
The results of antibiotics exposure are summarized in Table3. Total of 214 HAIs cases (91 cases before the diagnosis of HAIs, 123 cases after the diagnosis of HAIs) were exposed to antibiotic, compared with 157 of the non-HAIs controls. No signi cant difference in antibiotics exposure was found between HAIs cases and non-HAIs controls. In the cases, the percentage of patients with antibiotic combination therapy (antibiotic combination therapy, 33.3%) was higher than that of the controls (antibiotic combination therapy, 8.9%) (P<0.001). The OR for antibiotic combinations exposure in cases compared with non-HAIs controls was 4.92 (95% CI3.17-7.63). The types of antibiotics and route of drug administration were similar for case patients and controls (P>0.05).
The adjusted odds ratios of HAIs associated with antibiotics and antibiotic combination exposure are displayed in Table3. After multivariate analysis, the OR for antibiotics exposure in cases compared with the controls was 0.18 (95% CI 0.09-0.35), which might be a protective factor. In addition, the OR for antibiotic combination exposure in cases compared with the controls was 3.34(95% CI 1.48-7.51), which was a risk factor for HAIs. At the same time, we also found that the odds ratio increases with age and length of residence.

Discussion
In univariate analysis, we found no statistically signi cant association between HAIs and exposure to antibiotics. However, after adjusting for confounding factors such as age and length of stay, use of antibiotics before the diagnosis of hospital infection may be associated with HAIs, and may be a protective factor. In addition, the combined use of antibiotics may increase the risk of hospital infections. This is a novel and important discovery, as antibiotics are widely overused [10,18], HAIs are the main threat, and conventional control measures will require an unprecedented level of global cooperation.
Although prior studies have examined the relationship between HAIs and exposure to antibiotics, however, no study to date has systematically investigated the association between antibiotic exposure and HAIs. Thus, these previous results are questionable. Our discovery contrast with those of certain observational studies that have suggested antibiotic exposure is a risk factor for HAIs [19][20][21][22]. However, these results should be compared with caution because the infection sites and bacterial species in these studies are different. To date, datas from studies speci cally focusing on risk factors of different types of hospital infections have been con icting. A case-control study in a 600-bed tertiary-care teaching hospital setting demonstrated that no cephalosporin class was independently associated with ESBLs BSI; however, in a secondary model considering all oxyimino-cephalosporins as a single variable, a signi cant association was demonstrated [19]. Additionally, a retrospective observational cohort study conducted at the Hospital de Clínicas de Porto Alegre showed that the use of antibiotics within the last 10 days before the diagnosis of hospital-acquired pneumonia was the only independent predictor of infection with multidrug-resistant bacteria [21]. Also a recent study showing that prior use of antibiotics is the main factor for the presence of infection healthcare-associated endocarditis [23]. In addition, some studies did not clarify the relationship between hospital infection and antibiotics exposure [13][14][15]. Disagreement in the ndings may be due to differences in type of HAIs, adjusted covariates, patient population, and confounding by determination of antibiotic exposure time.
After multivariate analysis, the confounding factors of age, hospitalization and invasive procedures were adjusted, indicating that exposure to antibiotics is a protective factor for HAIs. Consider that if a patient uses antibiotics because of an infection in one part of the body, infections in other parts can be prevented, because the effect of antibiotics is usually systemic. However, unless the patient is very likely to be infected, it is not recommended to use antibiotics in advance to prevent unknown infections, and the abuse of antibiotics can lead to the production of resistant bacteria.
This study showed that use of the combination antibiotic was strongly associated with HAIs. There is few research reports on whether combination antibiotic could be a risk factor for acquisition of HAIs, and the type of hospital infection is single. The study conducted at the University of Maryland Medical Center found previous in-hospital piperacillin-tazobactam use (current admission), and previous present admission in-hospital vancomycin use (current admission) as independent risk factors [24]. A casecontrol study in Southern Brazil showed that use of the combination antibiotic piperacillin-tazobactam within the previous 14 days was strongly associated with extended-spectrum-blactamase production in bloodstream infections due to Klebsiella pneumonia or Escherichia coli [20]. According to another study, prior use of broad-spectrum drugs (third-generation cephalosporin, uoroquinolone, and/or imipenem) was a risk factor associated with ventilator-associated pneumonia caused by potentially drug-resistant bacteria (OR,4.12; 95% CI, 1. 2-14.2)[25]. In addition, a study conducted in China showed that the use of more than three antibacterial drugs in patients is an independent risk factor for piperacillin-tazobactam caused by carbapenem-resistant Klebsiella pneumonia [26]. Combining the above, whether it is a single infection type or a single antibiotic combination, and the results of this comprehensive study, the combination of antibiotics does increase the risk of HAIs. It is unclear why antibiotic combination is the risk of HAIs. One possible explanation is that the current state of antibiotic therapy has reached a critical point because indiscriminate usage is leading to both compromised immunity and increased resistance. Even the administration of combined broad-spectrum antibiotic therapy can lead to increased mortality in uninfected patients [27].
The major advantage of our study is that cases contain multiple types of infections, and the selection of controls is based on strati ed sampling in different quarters and matched by department. This study focuses on the relationship between antibiotic combination exposure and the health care-associated infections. In addition, it is different from the previous analysis of only some xed combination antibiotics. The focus of this study is the combination of different types of antibiotics. Furthermore, the diversi ed case mix is reassuring, and the results can be extended to similar centers, including patients of all ages and all educational backgrounds. All available records, including progress notes, prescription orders, medical advice and electronic records are reviewed. Inevitably, our study has several limitations. First, it was carried out in a single hospital, which may produce an impact on our results given variations in HAIs prevention and control practices among institutions. Regarding statistical analysis, what may threaten our results was confounding bias. For example, we couldn't align the cases and controls with underlying diseases, because the condition of patients with multiple underlying diseases was too complicated, although we tried to control this confounding bias through department matching. In addition, we were unable to identify any 'over-the-counter' use of antibiotics; however, 'over-the-counter ' antibiotics basically did not happen because antibiotics were strictly controlled drugs, so would not have affected the study cohort. Finally, the patient's exposure to antibiotics before admission was not considered because there was no reliable record.

Conclusions
In conclusion, this study suggests that exposure to combined antibiotics is a risk factor for HAIs, not indicate an increased risk of HAIs associated with exposure to antibiotics before occurrence of HAIs. As far as the author knows, this is the rst time this risk factor has been found in the population of a Chinese hospital. Applying evidence-based guidelines to the appropriate use of antibiotics may reduce the spread of bacteria without harming patients who really need this effective antibiotic treatment. However, excessive combined use of antibiotics may cause a series of harms, such as antibiotic resistance and the increased risk of hospital infections found in this study. Given this potential adverse effects, clinicians should use caution in prescribing combination antibiotic for patients at risk. Abbreviations health care-associated infections , HAIs.

Declarations
Ethics approval and consent to participate: The study was approved by the a liated hospital of Northwest Minzu University Research Ethics Committee.

Consent for publication:
Not applicable Availability of data and materials: All data generated or analysed during this study are included in this published article.

Competing interests:
The authors declare that they have no competing interests Funding: This research did not receive any speci c grant from funding agencies in the public, commercial, or notfor-pro t sectors.
Authors' contributions: Xiao-Liang Zhang analyzed and interpreted the patient data regarding the health care-associated infections and antibiotics exposure. Fang-Bin Li performed the datas collection, and was a major contributor in writing the manuscript. All authors read and approved the nal manuscript. c Two or more routes of administration.
Table3. The multiple logistic regression model adjusted the odds ratio of antibiotics and antibiotic combination exposure, and Identify variables associated with HAIs