建立和確效與cephamycin類抗生素相關出血事件之風險預測量表

Abstract

Background: Cephamycins are a group of unique cephalosporins in perspective of the substitution group at 7α of the cephem structure. The difference in chemical structure expanded their spectrum of activity and enabled cephamycins a popular choice among prescribers. With the surged amount of prescriptions, a growing number of suspected cephamycin-associated hemorrhagic events were reported, which raised the attention of the public. Several studies had proposed possible attributable mechanisms and attempted to identify risk factors for the adverse events. The diminished production of vitamin K by intestinal microflora resulted from the eradication by cephamycins and the inactivation of the coagulation factors due to impaired vitamin K cycle interfered by N-methylthiotetrazole (NMTT) leaving group of cephamycins are the two most convincible hypotheses for hemorrhagic outcomes to date. Other risk factors including malnutrition, advanced age, and renal/hepatic dysfunction were also suggested. Recent studies using National Health Insurance Research Database (NHIRD) and database of a teaching hospital in Taipei had tried identifying the risk factors, but the NHIRD was lack of laboratory information, and insufficient number of users of cephamycins were identified by the other study. The mentioned limitations urged us to conduct this research using National Taiwan University Hospital (NTUH) electronic medical records to determine risk factors for cephamycin-associated hemorrhagic events and further produce a prediction risk scoring system for clinical pharmacists to use and to enhance the quality of care as our ultimate goal. Methods: Adults at least 20 years old and consecutively used study antibiotics (cefmetazole, flomoxef, cefoxitin, cefoperazone as cephamycins; amoxicillin/clavulanic acid, ampicillin/sulbactam, cefuroxime, cefotaxime, ceftriaxone as reference) for more than 48 hours at the emergency/inpatient departments of NTUH during January 1st, 2009 to December 31st, 2015 were included for observation of any hemorrhagic events occurred during the following ten days since the initiation of study antibiotics (index date). The population was separated into two cohorts by the end of 2013 for evaluation of risk factors (2009-2013 cohorts) and validation of the established scoring system using the identified risk factors (2014-2015 cohort). Univariate logistic regression was adapted for assessment of the crude association between factors and the outcome of interest. Multivariate logistic regression was used for the adjusted association. Results of the multivariate logistic regression were treated as the foundation to develop the risk scoring system. Both cohorts contributed in the validation of the scoring system. Results: There were 46402 (contributed from 37094 individuals) and 22681 episodes (from 19201 individuals) identified at 2009-2013 and 2014-2015 cohorts, respectively. The 2009-2013 cohort had 18821 episodes of cephamycins and 27581 episodes of reference antibiotics, with 227 and 129 hemorrhagic events among respective study groups. The 2014-2015 cohort had 8615 episodes of cephamycins and 14066 episodes of reference antibiotics, with 116 and 88 hemorrhagic events among respective study groups. Gastrointestinal bleeding was the majority of the events across both cohorts (85.11% vs. 95.10%). Use of cephamycins was associated with a higher risk for hemorrhagic outcomes (aOR 2.03, 95% CI 1.60-2.58). Other risk factors including chronic hepatic disease (aOR 2.08, 95% CI 1.54-2.81), at least 65 years old (aOR 1.66, 95% CI 1.31-2.09), prominent bleeding tendency (aOR 2.46, 95% CI 1.94-3.12), and bleeding history within previous 180 days of the index date (aOR 6.84, 95% CI 5.38-8.68). A nine-score risk scoring system (AUROC=0.8035, 95% CI 0.7794-0.8275; Hosmer-Lemeshow goodness-of-fit test p=0.1044) named CHABB was developed based on the identified risk factors (1, 1, 1, 2, 4 points for each item), with higher scores indicating higher risk for bleeding. Conclusion: Use of cephamycins was associated with higher hemorrhagic events compared with commonly used penicillins and cephalosporins. The established risk scoring system, CHABB, may help clinical pharmacists identify high-risk patients, provide recommendations and monitor adverse events in cephamycins users according to the predictive risk for developing hemorrhagic events, and eventually enhancing the overall quality of care.