Computerized Tool Guides Antimicrobial Use

Major Finding: After implementation of a computerized antibiotic stewardship system, the rate of appropriate prescribing rose from 8% to 92% for tigecycline and from 19% to 91% for linezolid.

Data Source: A study conducted during two 4-month periods at a 214-bed hospital.

Disclosures: Dr. Po reported that he had no disclosures.

ATLANTA — A computerized antibiotic stewardship system for tigecycline and linezolid led to an overall decrease in prescriptions of both medications and a significant increase in their appropriate use in a 214-bed community hospital.

A physician order entry system with decision support “provides a nonconfrontational, evidence-based system that can be rapidly implemented,” Dr. John Leander Po and his associates said in a poster at the conference.

Previous strategies to reduce unnecessary use of antibiotics have included prior authorization, prescriber feedback/education, and antibiotic order forms. But little is known about the effectiveness of a computerized interface that is triggered whenever a prescription is entered, requiring input, said Dr. Po and his associates, of Banner Estrella Medical Center, Phoenix.

The onscreen system used in this study was designed to limit utilization of tigecycline and linezolid outside of Food and Drug Administration–approved indications. The FDA has approved the drugs for treatment of patients infected with a multidrug-resistant organism with no other options, for those at risk of penicillin or vancomycin anaphylaxis with no other options, or as second-line therapy for pneumonia, urinary tract infection, and staphylococcal infection. The interface also delivered recommendations for alternative antibiotics, with hyperlinks to evidence-based articles.

Antimicrobial use was monitored, and direct feedback was delivered to the prescriber—primarily hospitalists, surgeons, and emergency physicians—when inappropriate use of either antimicrobial occurred.

During the 4 months before the intervention, 36 prescriptions for tigecycline were ordered, compared with 12 during the 4 months with the computerized system in place. The proportion of appropriate orders rose significantly, from 8% (3) to 92% (11). Examples of inappropriate use in the preintervention phase included for empiric postoperative prophylaxis; for gastroenteritis; when a single, narrow-spectrum antibiotic was indicated (i.e., vancomycin); and in a patient without penicillin allergy.

During the intervention, the one inappropriate tigecycline prescription was for a postcolectomy patient with fever and negative blood cultures and no evidence of penicillin allergy.

Similarly, total linezolid prescriptions fell from 168 to 3 with the computerized system, and the proportion of appropriate orders also increased significantly, from 19% to 91%. Inappropriate linezolid use prior to the intervention included empiric therapy for skin and soft tissue infection (SSTI) and initial therapy for methicillin-resistant Staphylococcus aureus bacteremia and endocarditis. After the intervention, inappropriate use included empiric SSTI therapy, initial therapy for osteomyelitis, and vancomycin-resistant urinary tract infection with a negative urinalysis, Dr. Po and his associates reported.

The computerized system used in this study could serve as a model to reduce inappropriate prescribing of other antimicrobial agents, they commented.

My Take

Getting to the Root Cause of Inappropriate Prescribing

The study by Dr. Po and his associates offers several insights that are relevant to hospitalists. Clearly, the findings speak to the need for improvement in antimicrobial stewardship. We're doing better than we have in the past, but we still need to think before we prescribe. The system used in this study is a simple method designed to catch us if we don't.

A system that walks the user through a clear algorithm can not only steer us toward evidence-based best practice but also provide education at the point of care. Yes, we all know about appropriate antimicrobial prescribing, but there is tremendous variation around the country. Better standardization of these practices could lead to improved care and resource utilization.

The literature supports the use of decision-support systems, particularly for busy clinicians who are working in high-volume institutions. Sometimes we get overwhelmed, and these computerized systems can help us opt into the best standard of care. That's also why checklists have received so much attention lately. They use a similar approach and have been shown to improve outcomes.

However, the idea that pop-up alerts are the answer to everything comes with a price. After a while, too many alerts can lead to “alert fatigue,” and you stop paying attention. Just as we need to adopt more judicious use of our medications, technology needs to be applied judiciously as well.

Downstream alerts are a useful safety-net solution that can help reduce practice deviations. But it is always important to understand and investigate the root cause of why deviations occur in the first place. It's important for clinicians to step back, figure out why this is happening, and address it at the outset.

Major Finding: After implementation of a computerized antibiotic stewardship system, the rate of appropriate prescribing rose from 8% to 92% for tigecycline and from 19% to 91% for linezolid.

Data Source: A study conducted during two 4-month periods at a 214-bed hospital.

Disclosures: Dr. Po reported that he had no disclosures.

ATLANTA — A computerized antibiotic stewardship system for tigecycline and linezolid led to an overall decrease in prescriptions of both medications and a significant increase in their appropriate use in a 214-bed community hospital.

A physician order entry system with decision support “provides a nonconfrontational, evidence-based system that can be rapidly implemented,” Dr. John Leander Po and his associates said in a poster at the conference.

Previous strategies to reduce unnecessary use of antibiotics have included prior authorization, prescriber feedback/education, and antibiotic order forms. But little is known about the effectiveness of a computerized interface that is triggered whenever a prescription is entered, requiring input, said Dr. Po and his associates, of Banner Estrella Medical Center, Phoenix.

The onscreen system used in this study was designed to limit utilization of tigecycline and linezolid outside of Food and Drug Administration–approved indications. The FDA has approved the drugs for treatment of patients infected with a multidrug-resistant organism with no other options, for those at risk of penicillin or vancomycin anaphylaxis with no other options, or as second-line therapy for pneumonia, urinary tract infection, and staphylococcal infection. The interface also delivered recommendations for alternative antibiotics, with hyperlinks to evidence-based articles.

Antimicrobial use was monitored, and direct feedback was delivered to the prescriber—primarily hospitalists, surgeons, and emergency physicians—when inappropriate use of either antimicrobial occurred.

During the 4 months before the intervention, 36 prescriptions for tigecycline were ordered, compared with 12 during the 4 months with the computerized system in place. The proportion of appropriate orders rose significantly, from 8% (3) to 92% (11). Examples of inappropriate use in the preintervention phase included for empiric postoperative prophylaxis; for gastroenteritis; when a single, narrow-spectrum antibiotic was indicated (i.e., vancomycin); and in a patient without penicillin allergy.

During the intervention, the one inappropriate tigecycline prescription was for a postcolectomy patient with fever and negative blood cultures and no evidence of penicillin allergy.

Similarly, total linezolid prescriptions fell from 168 to 3 with the computerized system, and the proportion of appropriate orders also increased significantly, from 19% to 91%. Inappropriate linezolid use prior to the intervention included empiric therapy for skin and soft tissue infection (SSTI) and initial therapy for methicillin-resistant Staphylococcus aureus bacteremia and endocarditis. After the intervention, inappropriate use included empiric SSTI therapy, initial therapy for osteomyelitis, and vancomycin-resistant urinary tract infection with a negative urinalysis, Dr. Po and his associates reported.

The computerized system used in this study could serve as a model to reduce inappropriate prescribing of other antimicrobial agents, they commented.

My Take

Getting to the Root Cause of Inappropriate Prescribing

The study by Dr. Po and his associates offers several insights that are relevant to hospitalists. Clearly, the findings speak to the need for improvement in antimicrobial stewardship. We're doing better than we have in the past, but we still need to think before we prescribe. The system used in this study is a simple method designed to catch us if we don't.

A system that walks the user through a clear algorithm can not only steer us toward evidence-based best practice but also provide education at the point of care. Yes, we all know about appropriate antimicrobial prescribing, but there is tremendous variation around the country. Better standardization of these practices could lead to improved care and resource utilization.

The literature supports the use of decision-support systems, particularly for busy clinicians who are working in high-volume institutions. Sometimes we get overwhelmed, and these computerized systems can help us opt into the best standard of care. That's also why checklists have received so much attention lately. They use a similar approach and have been shown to improve outcomes.

However, the idea that pop-up alerts are the answer to everything comes with a price. After a while, too many alerts can lead to “alert fatigue,” and you stop paying attention. Just as we need to adopt more judicious use of our medications, technology needs to be applied judiciously as well.

Downstream alerts are a useful safety-net solution that can help reduce practice deviations. But it is always important to understand and investigate the root cause of why deviations occur in the first place. It's important for clinicians to step back, figure out why this is happening, and address it at the outset.

Major Finding: After implementation of a computerized antibiotic stewardship system, the rate of appropriate prescribing rose from 8% to 92% for tigecycline and from 19% to 91% for linezolid.

Data Source: A study conducted during two 4-month periods at a 214-bed hospital.

Disclosures: Dr. Po reported that he had no disclosures.

ATLANTA — A computerized antibiotic stewardship system for tigecycline and linezolid led to an overall decrease in prescriptions of both medications and a significant increase in their appropriate use in a 214-bed community hospital.

A physician order entry system with decision support “provides a nonconfrontational, evidence-based system that can be rapidly implemented,” Dr. John Leander Po and his associates said in a poster at the conference.

Previous strategies to reduce unnecessary use of antibiotics have included prior authorization, prescriber feedback/education, and antibiotic order forms. But little is known about the effectiveness of a computerized interface that is triggered whenever a prescription is entered, requiring input, said Dr. Po and his associates, of Banner Estrella Medical Center, Phoenix.

The onscreen system used in this study was designed to limit utilization of tigecycline and linezolid outside of Food and Drug Administration–approved indications. The FDA has approved the drugs for treatment of patients infected with a multidrug-resistant organism with no other options, for those at risk of penicillin or vancomycin anaphylaxis with no other options, or as second-line therapy for pneumonia, urinary tract infection, and staphylococcal infection. The interface also delivered recommendations for alternative antibiotics, with hyperlinks to evidence-based articles.

Antimicrobial use was monitored, and direct feedback was delivered to the prescriber—primarily hospitalists, surgeons, and emergency physicians—when inappropriate use of either antimicrobial occurred.

During the 4 months before the intervention, 36 prescriptions for tigecycline were ordered, compared with 12 during the 4 months with the computerized system in place. The proportion of appropriate orders rose significantly, from 8% (3) to 92% (11). Examples of inappropriate use in the preintervention phase included for empiric postoperative prophylaxis; for gastroenteritis; when a single, narrow-spectrum antibiotic was indicated (i.e., vancomycin); and in a patient without penicillin allergy.

During the intervention, the one inappropriate tigecycline prescription was for a postcolectomy patient with fever and negative blood cultures and no evidence of penicillin allergy.

Similarly, total linezolid prescriptions fell from 168 to 3 with the computerized system, and the proportion of appropriate orders also increased significantly, from 19% to 91%. Inappropriate linezolid use prior to the intervention included empiric therapy for skin and soft tissue infection (SSTI) and initial therapy for methicillin-resistant Staphylococcus aureus bacteremia and endocarditis. After the intervention, inappropriate use included empiric SSTI therapy, initial therapy for osteomyelitis, and vancomycin-resistant urinary tract infection with a negative urinalysis, Dr. Po and his associates reported.

The computerized system used in this study could serve as a model to reduce inappropriate prescribing of other antimicrobial agents, they commented.

My Take

Getting to the Root Cause of Inappropriate Prescribing

The study by Dr. Po and his associates offers several insights that are relevant to hospitalists. Clearly, the findings speak to the need for improvement in antimicrobial stewardship. We're doing better than we have in the past, but we still need to think before we prescribe. The system used in this study is a simple method designed to catch us if we don't.

A system that walks the user through a clear algorithm can not only steer us toward evidence-based best practice but also provide education at the point of care. Yes, we all know about appropriate antimicrobial prescribing, but there is tremendous variation around the country. Better standardization of these practices could lead to improved care and resource utilization.

The literature supports the use of decision-support systems, particularly for busy clinicians who are working in high-volume institutions. Sometimes we get overwhelmed, and these computerized systems can help us opt into the best standard of care. That's also why checklists have received so much attention lately. They use a similar approach and have been shown to improve outcomes.

However, the idea that pop-up alerts are the answer to everything comes with a price. After a while, too many alerts can lead to “alert fatigue,” and you stop paying attention. Just as we need to adopt more judicious use of our medications, technology needs to be applied judiciously as well.

Downstream alerts are a useful safety-net solution that can help reduce practice deviations. But it is always important to understand and investigate the root cause of why deviations occur in the first place. It's important for clinicians to step back, figure out why this is happening, and address it at the outset.