The Case

A 65-year-old woman was admitted to the intensive care unit (ICU) with severe sepsis and respiratory failure secondary to community-acquired pneumonia. The patient was intubated and treated with broad-spectrum antibiotics while waiting for blood culture results. A few hours into the admission, the microbiology lab called the ICU to report that the blood cultures were growing yeast. The report was received by the nurse and relayed to the covering resident. The patient was immediately started on treatment with intravenous (IV) amphotericin (an antifungal drug), and broad-spectrum antibiotics were continued.

Two days later, the patient's respiratory status was improving and she was extubated. However, she was noted to have worsening renal function and decreased urine output, and she needed to be placed on hemodialysis on hospital day 3. The consulting nephrologist felt that amphotericin toxicity was the most likely cause of the acute kidney injury. A new resident took over the patient's care that day, after the previous resident rotated off service. She reviewed the laboratory data and was surprised to find that the patient had no positive blood cultures reported in the electronic medical record. She called the microbiology lab, who confirmed that the patient had not had any positive blood cultures. In the course of the discussion, she came to a troubling realization: the positive yeast cultures were actually for the patient in the adjacent ICU bed.

Further review revealed that the information about the positive yeast culture had been correctly called in to the bedside nurse, who then relayed to the correct resident. However, the resident had been caring for both patients, and he mistakenly entered the amphotericin order for the wrong patient.

At this hospital, prescribing IV amphotericin required approval from an infectious disease specialist due to its potential for toxicity. The resident who placed the order had, in fact, called the infectious disease fellow, described the correct patient, and obtained approval for the medication. However, once the verbal approval was secured, no further verification was required—the resident merely had to enter "infectious disease approval obtained" into the free-text section of the electronic order for amphotericin.

Once the error was detected, the amphotericin was immediately stopped. The patient who had received the drug continued to require hemodialysis for several more days. Although she was not dialysis-dependent when discharged from the hospital, her kidney function never returned to her previous baseline. The other patient—who should have been started on amphotericin—did not experience any adverse consequences as a result of the delay in initiating appropriate therapy.

The Commentary

by Gregory A. Filice, MD

Before addressing the elephant in the room—the prescription of a highly toxic medication the wrong patient—it's worth noting that the first error in this case was the selection of amphotericin for treatment. The great majority of yeast infections in this setting are with Candida species. For treatment of candidiasis, echinocandins are more effective and associated with less severe toxicity than amphotericin.(1,2) There is moderate quality evidence that a clinical decision support system (CDSS) can improve antimicrobial selection and dosing and can reduce adverse events, including kidney damage, mortality, and length of stay.(3-5) Had there been a decision support system for antimicrobial use at the point of order entry, the resident might have selected an echinocandin for therapy instead of amphotericin.

In this hospital, an order for intravenous amphotericin required approval from an infectious disease (ID) specialist. A drug approval should be viewed as an opportunity for a very brief consultation. Unless the patient (Patient A) had risk factors for another infection that can present with yeast in the blood (i.e., histoplasma), the ID specialist should have recommended an echinocandin. Let's turn now to the wrong-patient error, which occurred when the resident wrote the amphotericin order for Patient B instead of Patient A. Even with widespread use of computerized provider order entry, drug orders for the wrong patient still occur with surprising frequency.(6,7) There were several opportunities for early recognition of the errors in this case.

Most hospitals with a restricted antimicrobial formulary require that an ID specialist communicate an antimicrobial approval directly to the pharmacy along with patient identifiers, the drug, and route. In this case, the approval would have been registered for Patient A and not for Patient B, and the error would have been detected before the drug was administered to Patient B. The pharmacy was willing to accept the provider's word for the approval—a process that allows for errors, misunderstandings, and even prescriber attempts to circumvent the approval system.(8)

The value of direct contact between the ID specialist and the laboratory goes beyond simple logistics—there is evidence that such communication can improve care.(9) Some hospitals make significant positive microbiology results available to ID specialists daily, and the ID specialists then review the electronic health record (EHR) for patients with these results to determine whether additional diagnostic or therapeutic measures are needed. If so, the ID specialists contact providers for these cases. In this case, an ID specialist reviewing significant microbiology results would likely have recognized that Patient A with yeast in a blood culture was not on appropriate therapy and would have contacted the provider.

The wrong-patient error was also not detected by the clinicians in the ICU. One might have hoped that the ICU physicians would have noticed the absence of antifungal therapy for Patient A or the unexpected antifungal therapy for Patient B. Optimal ICU care occurs when responsibility for each patient is provided by a team including one or more physicians, nurses, and pharmacists.(10) Team members are aware of key aspects of each patient's medical problems and laboratory reports. Teamwork is strongly reinforced by daily interdisciplinary rounds. Typically, team members go over the clinical details for each patient every day. A checklist of items, including a review of each patient's medications, is assessed daily with prompting.(11) Multidisciplinary ICU team rounds improve care.(12) Had this care model been in place, it is likely that an ICU nurse or pharmacist would have recognized the absence of therapy for the yeast infection in Patient A and the odd inclusion of antifungal therapy for community-acquired pneumonia in Patient B.

The majority of hospitals in the United States have antimicrobial stewardship programs (13), but components of the programs vary. Preauthorization and prospective audit and feedback improve antibiotic use, and one or both should be a core component of a stewardship program.(4) In a typical prospective audit and feedback program, inpatient antimicrobial administration is reviewed daily. In this case, a strong audit and feedback team would likely have recognized that there was no indication for amphotericin for Patient B and communicated this to the patient's provider.

The other activities of antimicrobial stewardship that bear on this case are didactic education and use of the EHR to support stewardship. Didactic education might have led the resident or the ID specialist to use an echinocandin instead of amphotericin, but there is little evidence that didactic education improves care.(4) EHRs and CDSSs have the potential to enhance antimicrobial stewardship in many ways. EHRs can be used to identify cases in which potential pathogens are identified and yet antimicrobial therapy is not given for these pathogens or is not likely to be effective ("bug–drug mismatches").(5,14) Evidence regarding the impact of EHRs on patient outcomes through improved antimicrobial use is lacking, although EHRs with integrated CDSSs have demonstrated improvements in clinical and economic outcomes.

There are several challenges to use of the EHR to support stewardship. Most EHRs do not have decision support for antimicrobial use in the basic package. Some major EHR vendors are developing the capacity to support antimicrobial stewardship, but this capacity will probably come at an extra cost.(5) Creation of a CDSS within existing EHRs requires many hours to develop, build, and test to ensure that it is both functional and efficient. Standalone EHRs are available, but extensive time and expense are necessary for interface development and data stream validation. Published experience on use of an EHR to support stewardship has been mixed. For example, one study found that a CDSS has saved staff time and costs (15), but another determined that approximately 70% of all electronic alerts to stewardship personnel were not actionable.(16) Obstacles to using EHR to support stewardship include costs for equipment and software; personnel costs; and administrative, legal, and technical issues.(5) This is an area of rapid development, and I am hopeful that the effectiveness of these will be demonstrated and that obstacles will be reduced in the next few years.

The patient safety movement has taught us that all manner of errors—including entering errors on the wrong patient—are likely to occur in human systems. Hospitals must work to ensure that all providers are aware of common errors and take steps to avoid them. However, we have also learned that the goal of minimizing harm to patients cannot be reached through education alone. The best overall approach to optimal patient safety is for hospitals to establish and support a culture of continuous improvement and patient safety that involves and is embraced by all health care workers.(17) A vibrant culture of continuous improvement and teamwork would likely help providers avoid errors like the one in this case or recognize them immediately and correct them.

Take-Home Points

• An antibiotic approval call should be viewed as a mini-consult, and the infectious disease provider should expect to know enough about a case to give sound advice on the need for an antimicrobial and the optimal drug(s).
• All hospitals should have antimicrobial stewardship programs. Guidelines recommend that preauthorization or prospective audit and feedback be a core component.
• Multidisciplinary ICU care teams and daily ICU team rounds improve care.
• Moderate quality evidence indicates that clinical decision support systems can improve antibiotic selection and reduce adverse events associated with antibiotic use in hospitalized patients.
• Review of significant positive microbiology results by infectious disease specialists can detect and correct errors and improve care.

Gregory A. Filice, MD Chief, Infectious Disease Section, Veterans Affairs Medical Professor of Medicine University of Minnesota

References

1. Pappas PG, Kauffman CA, Andes DR, et al. Executive summary: clinical practice guideline for the management of candidiasis: 2016 update by the Infectious Diseases Society of America. Clin Infect Dis. 2016;62:409-417. [go to PubMed]

2. Andes DR, Safdar N, Baddley JW, et al. Impact of treatment strategy on outcomes in patients with candidemia and other forms of invasive candidiasis: a patient-level quantitative review of randomized trials. Clin Infect Dis. 2012;54:1110-1122. [go to PubMed]

3. Filice GA, Drekonja DM, Thurn JR, et al. Use of a computer decision support system and antimicrobial therapy appropriateness. Infect Control Hosp Epidemiol. 2013;34:558-565. [go to PubMed]

4. Barlam TF, Cosgrove SE, Abbo LM, et al. Implementing an antibiotic stewardship program: guidelines by the Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America. Clin Infect Dis. 2016;62:e51-e77. [go to PubMed]

5. Forrest GN, Van Schooneveld TC, Kullar R, Schulz LT, Duong P, Postelnick M. Use of electronic health records and clinical decision support systems for antimicrobial stewardship. Clin Infect Dis. 2014;59(suppl 3):S122-S133. [go to PubMed]

6. Henneman PL, Fisher DL, Henneman EA, et al. Providers do not verify patient identity during computer order entry. Acad Emerg Med. 2008;15:641-648. [go to PubMed]

7. Adelman JS, Kalkut GE, Schechter CB, et al. Understanding and preventing wrong-patient electronic orders: a randomized controlled trial. J Am Med Inform Assoc. 2013;20:305-310. [go to PubMed]

8. LaRosa LA, Fishman NO, Lautenbach E, Koppel RJ, Morales KH, Linkin DR. Evaluation of antimicrobial therapy orders circumventing an antimicrobial stewardship program: investigating the strategy of "stealth dosing." Infect Control Hosp Epidemiol. 2007;28:551-556. [go to PubMed]

9. Huang RSP, Guervil DJ, Hunter RL, Wanger A. Lower antibiotic costs attributable to clinical microbiology rounds. Diagn Microbiol Infect Dis. 2015;83:68-73. [go to PubMed]

10. Dietz AS, Pronovost PJ, Mendez-Tellez PA, et al. A systematic review of teamwork in the intensive care unit: what do we know about teamwork, team tasks, and improvement strategies? J Crit Care. 2014;29:908-914. [go to PubMed]

11. Weiss CH, Moazed F, McEvoy CA, et al. Prompting physicians to address a daily checklist and process of care and clinical outcomes: a single-site study. Am J Respir Crit Care Med. 2011;184:680-686. [go to PubMed]

12. Kim MM, Barnato AE, Angus DC, Fleisher LA, Kahn JM. The effect of multidisciplinary care teams on intensive care unit mortality. Arch Intern Med. 2010;170:369-376. [go to PubMed]

13. Pogorzelska-Maziarz M, Herzig CTA, Larson EL, Furuya EY, Perencevich EN, Stone PW. Implementation of antimicrobial stewardship policies in U.S. hospitals: findings from a national survey. Infect Control Hosp Epidemiol. 2015;36:261-264. [go to PubMed]

14. Antworth A, Collins CD, Kunapuli A, et al. Impact of an antimicrobial stewardship program comprehensive care bundle on management of candidemia. Pharmacotherapy. 2013;33:137-143. [go to PubMed]

15. McGregor JC, Weekes E, Forrest GN, et al. Impact of a computerized clinical decision support system on reducing inappropriate antimicrobial use: a randomized controlled trial. J Am Med Inform Assoc. 2006;13:378-384. [go to PubMed]

16. Hermsen ED, Van Schooneveld TC, Sayles H, Rupp ME. Implementation of a clinical decision support system for antimicrobial stewardship. Infect Control Hosp Epidemiol. 2012;33:412-415. [go to PubMed]

17. McFadden KL, Stock GN, Gowen CR III. Leadership, safety climate, and continuous quality improvement: impact on process quality and patient safety. Health Care Manage Rev. 2015;40:24-34. [go to PubMed]