The overuse and misuse of antimicrobial agents is a global problem that has led to the development of antimicrobial resistance in both the hospital and community setting. One of the primary strategies for combating resistance is through the use of antimicrobial stewardship programs (ASPs) [1
]. The ASP’s primary goal includes optimizing antimicrobial use through reduction of unnecessary antimicrobial use and confirmation of proper antimicrobial use (including drug, dose, route, and duration), in order to achieve the best clinical outcomes. While ASPs have been proven effective in both academic hospitals [3
] and smaller community hospitals [4
], small community hospitals tend to have more difficulty establishing these programs. These difficulties include staffing constraints, lack of funding, and lack of administrative and medical staff support [5
Vidant Medical Center (VMC) has had a successful ASP in place since 2001. The ASP uses a primary strategy of prospective audit with feedback. Given the success of the ASP at the tertiary care center, we expanded the ASP to six of the seven community hospitals within the Vidant Health (VH) system. In December 2011 the ASP was implemented at Vidant Roanoke-Chowan Hospital (VROA), in March 2012 at Vidant Bertie Hospital (VBER), Vidant Chowan Hospital (VCHO), and The Outer Banks Hospital (OBH), in October 2013 at Vidant Duplin Hospital (VDUP), and in December 2013 at Vidant Beaufort Hospital (VBEA). We were able to accomplish this process through use of the electronic medical record (EMR) Epic (Madison, WI, USA), which is shared across the VH system [6
]. To date, we are unable to locate any previous attempt at managing an ASP via EMR and central monitoring. We currently collect data on intervention outcomes, cost savings, physician acceptance rates, number of charts reviewed, number of recommendations made, anti-methicillin-resistant Staphylococcus aureus
(MRSA) drug use, anti-pseudomonal drug use, broad spectrum drug use, and total antimicrobial drug use.
In 2007, the IDSA and the Society for Healthcare Epidemiology of America (SHEA) released guidelines for developing institutional programs to better antimicrobial stewardship through use of the EMR [9
]. VH has demonstrated long-term beneficial effects of an ASP and has used the EMR as a means of optimizing antimicrobial use [6
]. This study is unique in that we can find no record of any hospital system using their EMR to remotely practice antimicrobial stewardship at community hospitals.
Following expansion of the ASP to the community hospitals, 40%–63% of charts reviewed resulted in a recommendation being made with an 81%–95% physician acceptance rate. None of these recommendations occurred before the ASP was extended remotely to the community hospitals. This method provides an option for antimicrobial stewardship for smaller community hospitals and shows that physicians are willing to accept ASPs remotely.
The most commonly accepted intervention noted is drug discontinuation. This was associated with an average antimicrobial drug cost savings of $20,860.25 per hospital for the 4 largest hospitals over an 18 month period from January 2014 through June 2015. Again, this does not take into account additional cost savings that occur when adverse events are avoided, drugs are changed IV to PO, patient outcomes are optimized, and antimicrobial resistance is avoided.
One of the main targets with each remote ASP was the reduction of quinolone use due to its increased risk for both Clostridium difficile
infections (CDI) and MRSA infections [10
]. Overall, we saw statistically significant decreases in quinolone use at hospitals B, D and E but did not see significant changes at hospitals A, C or F. This decrease in quinolone use may be a driving factor for the decrease in anti-pseudomonal drug usage as well. Two possible reasons why hospitals A and C did not experience decreases in quinolone use could be due to the fact that their quinolone use was considerably lower at the beginning of ASP implementation and because the ASP is newer at both of these hospitals.
Some hospitals did see an increase in cephalosporin use. This may be a result of implementation of a dose optimization protocol that attempted to maximize pharmacokinetic and pharmacodynamic properties for certain pathogens and patient populations. For example, all surgical cefazolin dosing was increased from 1 g to 2 g, empiric cefepime dosing for hospital acquired infections was increased from 1 g to 2 g every 12 h to 2 g every 8 h given by extended infusion, and ceftriaxone dosing was increased from 1 g to 2 g based on type of infection and patient specific parameters. While macrolide usage varied based on hospital, periodic analysis showed that most use of azithromycin is driven by the emergency department in the form of empiric sexually transmitted disease treatment or first dose for those not admitted to the hospital.
One important note to make is related to total antimicrobial use at each hospital. Overall, there was not a statistically significant decrease in total antimicrobial use at any of the community hospitals. The ASP does not review patients in the emergency department, those who come to the hospital daily for infusions, or those who are on antibiotics for less than the 24 h period it takes to flag on the report. However, all of this antimicrobial usage is included within the total usage reported. In addition, total usage includes antimicrobials that are not on the controlled list and that would never flag for ASP review.
The goal of ASPs is not only to reduce unnecessary use of antimicrobials, but also to improve resistance profiles. Because isolate numbers at each hospital were small, only hospital D’s isolate pool was large enough to analyze. There were improvements in antibiotic susceptibilities of P. aeruginosa to ciprofloxacin, piperacillin/tazobactam, and carbapenems over the four year time period.
Establishing a remote ASP is not without challenges. There can be variation in local resources including diagnostics and formulary. While the VH formulary is now standardized, there is still variation in what drugs are stocked by each pharmacy and there are currently no formulary restrictions at the community hospitals. Determining how to identify patients can also be a challenge and may have to be modified over time. Distinguishing cases that need stewardship assistance vs. a formal infectious diseases consult can also be a challenge.
Being a successful remote ASP does not stop with patient chart review. Continuing to develop relationships with the local staff (physicians, pharmacists, microbiology staff, and infection control practitioners) at the community hospitals is critical to improving patient care, as we view this as a team effort towards antimicrobial stewardship. The ASP pharmacists attempt to visit each community hospital on a yearly basis in order to provide some face-to-face interaction, conduct educational opportunities desired by the pharmacy or physician staff, share results of the program, and gather feedback. This process also allows formal ASP introduction to any new or temporary staff. It is common for acceptance rates of new physicians to be low until they become comfortable with the advantages of the program. In addition to daily chart review, the ASP has been responsible for tasks including, but not limited to, helping manage antimicrobial shortages and formulary, creating order sets, answering questions for the local wound care centers, and distributing a guide book that is updated yearly and includes key information about managing infectious diseases.
There are several limitations to this study. First, this study is based on aggregate data; the impact of the duration of antimicrobial use for an individual patient cannot be determined. Second, this dataset cannot correct for seasonal variation. Each ASP was implemented at a different time, with the oldest program running for four years and the youngest running for only one year. Because of these limitations, there were currently not enough data points to properly analyze antimicrobial patterns over the course of a year. Third, because of the small hospital sizes and small number of bacterial isolates, there was limited data regarding improvements in hospital antibiograms.