A 60-year-old woman with squamous cell carcinoma of the glottis underwent laryngectomy, anterior neck dissection, and pectoralis flap reconstruction of the anterior esophagus. Postoperatively, she was started on clindamycin for surgical site infection prophylaxis. On postoperative day three, increased drainage from the surgical site was noted, which progressed and required a return to the operating room for exploration on postoperative day six. The site revealed purulent material with involvement of the jugular sheath. Cultures of the wound were taken, and the site was irrigated and re-closed.
Two days later, the wound cultures grew Staphylococcus aureus. Per hospital protocol, the microbiology laboratory called the nursing unit to report the positive culture, and the patient's nurse informed the physician team. Clindamycin was continued. Three days later, a final sensitivity profile for the S. aureus isolate returned, showing resistance to clindamycin. Hospital policy does not call for notification to clinicians by the microbiology lab when a "preliminary" culture result becomes "final" (when final sensitivities become available). At this facility, physicians view laboratory results in a scrolling text–based computer system, which presents results in chronological order according to the time of specimen collection. The final microbiology result and the critical sensitivity results in this case remained in chronological order (now 5 days prior). The physicians did not view the updated culture results, and the patient remained on clindamycin.
The next day, the right internal jugular vein ruptured, and the patient was taken emergently to the operating room. A widespread infection of the surgical site with destruction of the jugular sheath and erosion into the internal jugular vein was noted. The internal jugular vein was ligated and sacrificed. The patient died of sepsis and multiorgan system failure 3 days later.
The case presented describes a 60-year-old woman who developed a serious postoperative surgical wound infection following laryngectomy, anterior neck dissection, and pectoralis flap reconstruction. The scope of potentially preventable medical errors ranges from inappropriate timing of surgical antibiotic prophylaxis to inaccurate selection of empiric antibiotics. While the procedures involved in reporting clinical microbiology laboratory results can create opportunities for error, they likely had little bearing on the outcome of this particular case, although the issue warrants some discussion.
Timing and Clinical Decision-Making
Importantly, the patient in this case did not receive preoperative antibiotics prior to the surgical procedure. The benefit of antibiotic prophylaxis in reducing surgical wound infection after major, clean-contaminated otolaryngological tumor resection is well-established.(1) The most commonly implicated pathogens associated with postoperative wound infection include normal oropharyngeal and nasal flora, including anaerobes, but also streptococci and staphylococci. In these elective procedures, clindamycin is active against these pathogens and is an appropriate choice for prophylaxis.(2) While clindamycin was an appropriate choice, prophylactic antibiotics must be administered immediately prior to surgical incision for maximal benefit. In contrast, administering postoperative antibiotics alone is no better than not using perioperative antibiotics at all.(3)
In our case, antibiotics were administered only after completion of the surgical procedure, a clear error. What measures can improve the proper administration of antibiotic prophylaxis? Obstacles to proper timing of surgical antibiotic prophylaxis include the perception that timing of administration is a "low priority," is inconvenient, and does not fit into the usual preoperative workflow. In addition, insufficient organizational communication and poor clarity of responsibility also have been identified as barriers.(4) A number of studies have identified mechanisms to improve the timing of preoperative antibiotics.(5,6) The most impressive results were seen from a multidisciplinary computer-guided support system, which was associated with improved timing of preoperative antibiotics and reduction in the rate of clean wound infection.(6) In this model, a multidisciplinary team reviewed published guidelines, developed antibiotic recommendations, and shifted the responsibility for administration from the preoperative nursing staff to the anesthesiologist. Appropriate timing of antibiotic administration improved from 51% to 98%, and clean wound infection rate decreased from 2.7% to 1.4%.
Moving back to this case, increased drainage from the surgical site was noted 3 days postoperatively while the patient received clindamycin. On the sixth postoperative day, the patient required re-exploration, still receiving empiric clindamycin. Considering the apparent lack of response to 6 days of clindamycin, an empiric broadening of antibiotic coverage to include agents more predictably active against methicillin-resistant Staphylococcus aureus (MRSA) and aerobic gram-negative bacilli should have occurred. Ultimately, the chosen empiric antibiotic did not match the susceptibility of the bacterial pathogen. Potential mechanisms to improve the selection of empiric antibiotics have centered upon computerized decision support systems.(7-9) In one study, a decision support system was found to reduce antibiotic-susceptibility mismatches associated with empiric antibiotics.(7) Similarly, a computerized antimicrobial guidance program based upon individual hospital clinical microbiology susceptibility data has been demonstrated to significantly improve the correct choice of empiric antibiotics associated with bacteremia.(8)
In this case, the initial clinical microbiology laboratory reported the growth of S. aureus 2 days after surgical re-exploration. Growth of S. aureus while the patient was on clindamycin should have alerted the practitioner to the possibility of clindamycin-resistant S. aureus and led to a change to empiric vancomycin. Ultimately, 3 days later, the susceptibility results were finalized, and the organism was determined to be resistant to clindamycin. The chronological order in which the susceptibility results were reported in this hospital might have impacted their review by the surgical team. However, considering that the rupture of the jugular vein and emergent return to the operating room took place the day after the susceptibility results were finalized, this staggered reporting system likely had little impact on the unfortunate outcome—particularly in comparison with the failure to initiate antibiotics preoperatively or empirically broaden them based on clinical response.
Staggered Reporting of Pathogens with Sensitivity Results
While the reporting system here may not have significantly influenced the outcome, a similar system may impact care more dramatically in different circumstances. Standard procedure calls for a clinical microbiology laboratory to provide an early report identifying the isolated pathogen, pending susceptibility results. The intent in providing early preliminary microbiology reports is to streamline empirical therapy to that consistent with the isolated pathogen. As mentioned previously, early identification of S. aureus in the highlighted case should have triggered a change in antibiotics, and such a change may have favorably affected the outcome. From a systems standpoint, standard practice also calls for the clinical laboratory to report microbiological data chronologically based upon the time of the sample, not at the time when the lab test was performed. One systems change that might increase the likelihood of a clinician viewing final susceptibility results from an earlier sampled culture might be an additional, updated clinical microbiology entry with finalized susceptibility results referring back to the original sample. However, the impact of such a reporting mechanism has not been confirmed in controlled studies, and two entries for one culture could result in confusion and error. Potentially, the best resolution to such reporting mismatch associated with a delay between identification of the pathogen and final susceptibility results might be the use of newer rapid susceptibility tests. In the case of MRSA, commercially available latex agglutination testing with associated identification of penicillin-binding protein 2a can confirm the presence of MRSA within 24 hours. Using this technology, the identification of the pathogen and susceptibility results would have been reported at the same time, eliminating the potential error associated with staggered reporting.
Finally, the use of integrated pharmacy systems—those with coordination between pharmacy and clinical microbiology databases—represents a potential future solution for pathogen-susceptibility mismatch. While some early commercial systems are currently marketed, the success of such systems remains to be proven. One system involves Therapeutic Antibiotic Monitoring (TAM) alerts in which coordinated pharmacy and microbiology databases notify the clinician of pathogen-antibacterial mismatch. However, one particular challenge with such systems is the identification of true mismatches as opposed to false positives (e.g., colonizers vs. true pathogens). As per the recently released Infectious Diseases Society of America and Society for Healthcare Epidemiology of America guidelines, the combination of coordinated pharmacy and microbiology databases along with an antibiotic stewardship team may be the optimal solution toward appropriate antibacterial use, including pathogen-antibacterial mismatch.(10)
- Appropriate timing of antibiotic prophylaxis and improved selection of empiric antibiotics are critical in the prevention of surgical site infections.
- Institution-specific programs can improve the quality of timing for antibiotic prophylaxis through multidisciplinary teams and clear roles and responsibility for timing of the preoperative dose.
- Systematic efforts to improve the selection of empiric antibiotics, through antimicrobial stewardship, include continuous summary, review, and dissemination of institutional susceptibility results with associated recommendations for therapy.
- Rapid susceptibility testing will minimize errors associated with staggered reporting of preliminary culture and final susceptibility testing.
B. Joseph Guglielmo, PharmD Professor and Thomas A. Oliver Chair in Clinical Pharmacy School of Pharmacy, University of California, San Francisco
1. ASHP Therapeutic Guidelines on Antimicrobial Prophylaxis in Surgery. Am J Health Sys Pharm. 1999;56:1839-1888. [go to PubMed]
2. Johnson JT, Yu VL, Myers EN, Wagner RL. An assessment of the need for gram-negative bacterial coverage in antibiotic prophylaxis for oncological head and neck surgery. J Infect Dis. 1987;155:331-333. [go to PubMed]
3. Classen DC, Evans RS, Pestotnick SL, Horn SD, Menlove RL, Burke JP. The timing of prophylactic administration of antibiotics and the risk of surgical-wound infection. N Engl J Med. 1992;326:281-286. [go to PubMed]
4. Tan JA, Naik VN, Lingard L. Exploring obstacles to proper timing of prophylactic antibiotics for surgical site infections. Qual Saf Health Care. 2006;15:32-38. [go to PubMed]
5. Burke JP. Maximizing appropriate antibiotic prophylaxis for surgical patients: an update from LDS Hospital, Salt Lake City. Clin Infect Dis. 2001;33(suppl 2):S78-S83. [go to PubMed]
6. Webb AL, Flagg RL, Fink AS. Reducing surgical site infections through a multidisciplinary computerized process for preoperative prophylactic antibiotic administration. Am J Surg. 2006;192:663-668. [go to PubMed]
7. Evans RS, Pestotnik SL, Classen DC, et al. A computer-assisted management program for antibiotics and other antiinfective agents. N Engl J Med. 1998;338:232-238. [go to PubMed]
8. Mullett CJ, Thomas JG, Smith CL, Sarwari AR, Khakoo RA. Computerized antimicrobial decision support: an offline evaluation of a database-driven empiric antimicrobial guidance program in hospitalized patients with a bloodstream infection. Int J Med Inform. 2004;73:455-460. [go to PubMed]
9. Davey P. The potential role of computerized decision support systems to improve empirical antibiotic prescribing. J Antimicrob Chemother. 2006;58:1105-1106. [go to PubMed]
10. Dellit TH, Owens RC, McGowan JE Jr, et al. Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America guidelines for developing an institutional program to enhance antimicrobial stewardship. Clin Infect Dis. 2007;44:159-177. [go to PubMed]