Clinical Decision Support and Appropriateness of Antimicrobial Prescribing – A Randomized Trial

From Clinfowiki
Jump to: navigation, search

This is a research article by Samore et al.,(2005). A three year(2001-2003) randomized trial was conducted to study the effect of CDSS in the appropriateness of antimicrobials prescription in rural community hospitals in ambulatory setting in Utah and Idaho States. titled "Clinical Decision Support and Appropriateness of Antimicrobial Prescribing – A Randomized Trial."[1]


To evaluate the influence of the Clinical Decision Support System (CDSS) Antimicrobial Prescribing in Community-based hospitals in ambulatory settings compared to Community prescribing patterns.[1]


To study the additive effect of antimicrobial prescribing with the help of combined CDSS and community to that of the only local/home community based prescribing patterns particularly rural areas for acute Upper Respiratory Infections (URI).[1]


The recent studies have documented a decrease in the antimicrobial prescribing for acute upper respiratory tract due to widespread notion that most URI’s are caused by viruses, except few acute URI like acute otitis media, strep throat. So, in order to have more stringent control over it, studies are essential especially in rural areas where the ratio of the high quality patient care to emergency care, availability to the world class facilities as in urban is significantly low.[1] Also the usage of antimicrobials is bestowned upon both clinicians and patients, as self-medication in patients is not controlled by clinician so also the overprescribing of antibiotics by clinicians in order to satisfy the patient’s desire of quick relief and recovery.

Study Design

Of the clustered randomly selected patients from 12 rural community hospitals under care by primary physicians in Utah and Idaho. From each state 2 large and 4 smaller community hospitals were all divided into 3 groups based on the intervention type and in last group 6 non-study communities were put as reference for controls to analyze retail pharmacy data. Also the study groups were studied in 3 equal time frames starting with prior intervention, first year post intervention and second year post intervention were January 2001 to September 2001, January 2002 to September 2002, and January 2003 to September 2003. First Group: Of the 6 random community hospitals with Community intervention alone, which was carried in two phases-one included the patient education to self-manage at home without antimicrobials, and another was to engage the clinician and patient communication. Second Group: These 6 random community hospitals,followed a combined CDSS intervention with Community intervention using 3 decision tools. These were two of paper-based versions and one on a handheld personal digital assistant(PDA). One paper version was a patient-initiated chart-documentation tool with specific symptoms options. The other paper version was an easy-to-use graphical flowchart. The PDA-based CDSS generated diagnostic and therapeutic recommendation with patient-specific information with the diagnosis based on symptoms and therapy with “Over the Counter" drugs and prescribed antimicrobials. Third non study group of 6 random hospitals were chosen to set as reference for retail pharmacy data. The two main sources of information for the three groups were retail pharmacy prescription with antimicrobials and chart review from primary care clinicians using CDSS.[1]


Diagnosis was done based on 4 most prescribed antimicrobials i.e. Penicillins, Macrolides, Cephalosporins and others were further categorized as never indicated” , “sometimes indicated” and “always indicated”. The trends in combined CDSS intervention with community intervention were compared with those of community intervention for the prescribed antibiotic class and in different time frames.[1]

Statistical analysis

The mixed (multilevel) regression models were applied to compare intervention arms with MLwiN 2.0 and Stata 8.0 Software programs. Parameter estimation: second order penalized quasi-likelihood. P values significance was less than .05.[1]


Exposure to intervention

The community intervention had reached almost 100% involvement by clinicians prescribing at the end of third year, in distribution OF self-care guides at different clinics, health fairs, and special community events. In combined CDSS and community clinician intervention, the highest percentage clinician's access was by hand held PDA (67%)followed by combination of PDA and paper tool, graphical flowchart on paper and patient initiated chart documentation.[1]

Trends in Antimicrobial Use

There was correlation between never indicated antimicrobial prescribing for diagnoses with the number of case-specific algorithms used. Within CDSS communities, clinicians who used algorithms of antimicrobial use and clinicians who did not use the tools. Both these correlations showed decreased trends. Within the CDSS intervention of all the antimicrobials the most macrolides prescribed drug azithromycin was decreased to 12% to 28% by the end of the study, followed by cephalosporins(7%) and penicillins(6%) respectively. There was no decline in prescribing of antimicrobials for diagnoses in the sometimes indicated or always indicated groups in either study group.[1]


  1. This study showed contradicting results to the current trends about the increased macrolide prescription in US and Europe.
  2. CDSS intervention was successful in keeping up the check of the overuse of antimicrobial prescriptions patterns.
  3. CDSS was proven to be successful in rural setting given the quality of care and it availability due to geographic separation of both clinicians and patients within their communities.[1]


  1. 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 Samore, M. H., Bateman,K., Alder S. C., Hannah, E., Donnelly S., Stoddard G. J., Haddadin B., Rubin, M. A., Williamson, J., Stults, B., Rupper, R., Stevenson K. The Journal of American Medical Association. 2005; 294(18):2305-2314. Retrieved from