Sepsis is the presence of an infection followed by a systemic inflammatory response (1, 2) and is responsible for more than one million hospitalizations per year within the United States (3). Severe sepsis is sepsis followed by organ dysfunction (2). The incidence of severe sepsis has been estimated to range between 300 and 1,031 per 100,000 per year, depending on ICD-9 (International Classification of Disease, Ninth Revision) coding (1, 4, 5). In 2007, severe sepsis was responsible for more than 200,000 deaths within the United States (3).
The current guidelines for treatment of sepsis generated by the Surviving Sepsis Campaign include administration of intravenous broad-spectrum antimicrobial therapy within the first hour of severe sepsis recognition (4). However, considerable differences of opinion exist as to which empiric therapies to start (6).
Previous studies have shown conflicting evidence regarding the impact of inappropriate initial antibiotic therapy (7–15). The purpose of this study was to evaluate the impact of inappropriate empiric antimicrobial therapy for the treatment of severe sepsis among a large cohort of surgical and trauma patients at a tertiary care center. We hypothesized that inappropriate initial antimicrobial therapy among patients would predict greater mortality risk than appropriate initial antimicrobial therapy.
MATERIALS AND METHODS
Institutional review board approval was obtained before data analysis. This was a retrospective analysis of a prospectively maintained database of all surgical patients (e.g., general, abdominal organ transplant, and trauma) admitted to the University of Virginia Health System from 1996 to 2007 and treated for sepsis (after 2007, study protocols dictating antibiotic timing and choice were enacted, such that the association between empiric therapy and outcome could no longer be independently determined). Data for each patient were prospectively collected every other day by chart review, patient interview/examination, and review of laboratory, microbiology, and pharmacy reports. Unique episodes of infection were identified for each patient and classified as separate if more than 72 h apart. Sites of infection, antibiotic therapy and duration, and organisms were recorded for each episode of infection. Patients were subsequently followed until death or hospital discharge. Initial antimicrobial therapy for each episode was compared with final culture results. Empiric therapy included all antibiotics started on the first day of treatment (administered shortly after blood cultures were taken), including those initiated for treatment of another simultaneous infection. End points considered for antibiotic treatment duration included culture-proven resistance or insensitivity, infection resolution, or death. Culture-specific antibiotics (definitive therapy) were subsequently initiated if resistance or insensitivity was determined.
Patient demographics and comorbidities evaluated at the time of initial infectious episode included gender, age, (patient-defined) race, solid organ transplant, trauma, diabetes mellitus, hypertension, hyperlipidemia, obesity, cardiovascular disease, peripheral vascular disease, pulmonary disease, ventilator dependence, renal insufficiency (RI), hemodialysis (HD) dependence, hepatic insufficiency, malignancy, chronic steroid use, human immunodeficiency virus (HIV), prior transfusion during same hospitalization, nosocomial infection, patient location at time of infection, Acute Physiology And Chronic Health Evaluation II (APACHE II) score, maximum temperature (Tmax), and white blood cell (WBC) count. Patient demographics and comorbidities measured at the time of each subsequent infectious episode included new-onset or change in ventilator dependence, RI, HD, transfusion, patient location, APACHE II score, Tmax, and WBC count. Similarly, sites of infection and organisms cultured were measured at the time of each infectious episode.
Infections were defined according to criteria specified by the Centers for Disease Control and Prevention (16). “Inappropriate” antimicrobial treatment was defined as empiric antibiotics initiated on day 1 of suspected sepsis that did not treat all organisms based on subsequent sensitivity testing compared with “appropriate” treatment defined as initial coverage that met these criteria (15, 17). The microbiology laboratory at our institution limits the number of isolated microbial species to three and provides quantification of all isolates, with the exception of blood cultures.
Fungal sensitivity testing was not performed during the study period because of availability reasons, and thus, Candida species were assumed sensitive to fluconazole, whereas Candida cruzii was considered resistant to fluconazole (18). Likewise, anaerobic sensitivity testing was not routinely performed, and thus Bacteroides fragilis and Bacteroides non-fragilis were assumed sensitive to flagyl, clindamycin, piperacillin/tazobactam, ampicillin/sulbactam, and all carbapenems (18). In addition, we assumed that all gram-positive cocci (GPC) were sensitive to vancomycin excluding vancomycin-resistant enterococci (VRE); we have no reported history of vancomycin-resistant Staphylococcus aureus or vancomycin-intermediate S. aureus at our institution (18). All GPC (as part of mixed infections) were assumed adequately treated by penicillins, cephalosporins, carbapenems, and fluoroquinolones unless proven otherwise by sensitivity testing (18). Finally, we assumed that all gram-negative rods were adequately treated by penicillins, cephalosporins, carbapenems, fluoroquinolones, aminoglycosides, and aztreonam unless proven otherwise by sensitivity testing (18).
“Solid organ transplant” was defined as kidney, liver, pancreas, heart, lung, kidney/pancreas, liver/pancreas, and small bowel. “Patient location at the time of infection episode” was defined as home, hospital ward, intensive care unit (ICU), or other. “Renal insufficiency” was defined as a serum creatinine of greater than or equal to 2.0 mg/dL at the time of infection episode. “Obesity” was defined as body mass index greater than 30. Pulmonary disease was defined as active treatment for lung disease before hospital admission. Other comorbidities were defined by chart review or patient examination. Mortality was defined as any cause of death after infection diagnosis while hospitalized.
Demographics and comorbidities, infection sites, infection organisms, and outcomes were compared between inappropriately and appropriately treated groups. Relative risk (RR), 95% confidence intervals (CIs), and P values were computed using a generalized estimating equation (GEE) approach with robust SEs (i.e., Huber White “sandwich variance” estimates) to accommodate for a correlated data structure corresponding to multiple episodes of infection per individual. The Deddens-Petersen Replication Method was used to compute point estimates and CIs when a convergence was not achieved with the traditional GEE model (19). Variables deemed statistically significant among the demographics and comorbidities, infection sites, and infection-related organisms were included in the multivariable log-binomial regression model. Analysis was performed using SAS Version 9.3 (Cary, NC) programming software. Statistical significance was defined as a value of P < 0.05.
A total of 2,855 patients with 7,158 separately identified episodes of infection were identified from our prospectively maintained database between 1996 and 2007. Infections were treated with empiric, antimicrobial therapy when sepsis was initially suspected. Empiric antimicrobial therapy was determined to be “inappropriate” for 597 patients (2,085 infectious episodes) and “appropriate” for 2,258 patients (5,073 infectious episodes) by culture results. There were 2.2 ± 2.4 episodes of infection per patient (inappropriate = 1.6 ± 1.3 vs. appropriate = 1.9 ± 1.7; P < 0.0001). Time from sepsis recognition to antibiotic initiation for each episode of infection was 0.59 ± 2.4 (inappropriate) vs. 0.67 ± 2.1 (appropriate) h (P = 0.66). Two hundred forty-one patients (1,035 infectious episodes) died during their hospitalization, whereas 2,614 patients (6,123 infectious episodes) lived.
Demographics and comorbidities stratified by appropriateness in empiric antimicrobial therapy are listed in Table 1. Age (older than 55 years), prior transfusion, nosocomial infection, APACHE II score (16 – 20), and WBC count (13.4 – 18.8) were more commonly associated with inappropriately treated episodes of infection compared with appropriately treated episodes of infection.
Sites of infection stratified by appropriateness in empiric antimicrobial therapy are listed in Table 2. Infections of the peritoneum, pleura, wound, and line were more commonly associated with inappropriately treated episodes of infection compared with appropriately treated episodes of infection.
Culture-proven organisms stratified by appropriateness in empiric antimicrobial therapy are listed in Table 3. Candida albicans, Candida glabrata, Enterobacter cloacae, methicillin-resistant S. aureus (MRSA), coagulase-negative Staphylococcus, Enterococcus faecalis, Enterococcus faecium, and VRE were more prevalent among inappropriately treated episodes of infection compared with appropriately treated episodes of infection.
Patient outcomes stratified by appropriateness in empiric antimicrobial therapy are listed in Table 4. Inappropriately treated patients received more antibiotics per episode of infection (P < 0.0001) for a longer duration (P < 0.0032) and experienced longer hospital length of stay (P = 0.0027) than appropriately treated patients. In addition, more inappropriately treated patients underwent surgical and or procedural source control compared with appropriately treated patients (P = 0.0008). However, a difference between groups was not observed regarding time from sepsis recognition to surgical/procedural source control (P = 0.23) or mortality (P = 0.19). After adjusting for statistically significant variables in Tables 1 to 3, inappropriate empiric antibiotic selection for treatment of suspected sepsis was not found to be associated with an increased risk for mortality compared with appropriate empiric antibiotic selection (RR, 1.0; 95% CI, 0.99 – 1.02; P = 0.36) (not included in Table).
Of the 7,158 infections treated empirically for sepsis at our hospital during the study period, 29% were determined to be inappropriate. This incidence is similar to that reported within the literature, varying between 9 and 56% (9, 10, 12–14). Of the inappropriately treated infections, approximately 317 (15%) resulted in death (51 patients). Our mortality incidence among inappropriately treated infections is similar to that reported within the literature, varying between 1.7 and 61.9% (12, 14, 20–22).
In contrast to previous studies citing a negative impact on survival (10–15), our study found that inappropriate empiric antibiotic therapy for treatment of suspected sepsis did not independently predict an increased risk for mortality among patients. Based on our sample size, the difference between our 95% upper and lower CI for our multivariable result was less than 0.03. This indicates that our study was sufficiently powered, and that the variance associated with the point estimate was extremely small and well within an equivalence region showing no difference from the null. Our results may be explained by the fact that a significantly greater number of infection episodes were present within the appropriately treated group but were also localized to the upper gastrointestinal tract and colon compared with the inappropriately treated group. Previous studies have observed an increased risk of mortality associated with intra-abdominal infections (12, 20–22). Turnbull et al. (23) evaluated genetically identical mice and their response to antibiotic therapy after receiving double-puncture cecal ligation. They observed an improvement in outcome with the incorporation of antibiotic use. However, there was a critical interleukin 6 threshold, above which mortality was assured regardless of antibiotic treatment. In our group of appropriately treated patients, a significantly fewer number of infectious episodes were managed surgically; walled-off abscesses or unresolved enteric leaks may be recalcitrant to antibiotics no matter how appropriate. Alternatively, inappropriately treated infections may have experienced improved survival regardless of inappropriateness because of surgical management and source control (10, 13, 14).
Many of the patients in our study experienced protracted hospital lengths of stay and were treated with multiple antibiotics for prolonged periods and for multiple concomitant infections; risk factors known to be associated with multidrug-resistant pathogens (24). Many of these organisms (e.g., MRSA and VRE) were more prevalent within the inappropriately treated group compared with the appropriately treated group. In addition, fungal infections (e.g., C. albicans and C. glabrata) were more prevalent within the inappropriately treated group compared with the appropriately treated group. This may be attributable to immunosuppression brought about by patient comorbidity or septic state or superinfection secondary to antibiotic use and thought to be caused by a disruption of the normal microflora, allowing opportunistic pathogens to proliferate (25). Both multidrug-resistant organisms and invasive fungal infections have previously been associated with poor outcomes, including death in critically ill patients (26–29). Recent attention has been devoted toward biomarker identification, metabolomic profiling, and/or rapid microbiologic microarrays to be used for the early diagnosis and prognosis of sepsis (29, 30). Further refinement, verification and validation, and incorporation of this technology may serve to expedite appropriate antibiotic selection before the onset of drug resistance.
Although the number of antibiotics and source control operations/procedures were significantly greater in the inappropriately treated group than the appropriately treated group, the time to antimicrobial therapy (inappropriate = 0.59 ± 2.4 h vs. appropriate = 0.67 ± 2.1 h; P = 0.66) and source control intervention (inappropriate = 19 ± 30 h vs. appropriate = 21 ± 30 h; P = 0.23) were similar between groups. As previously mentioned, current recommendations for the management of sepsis include broad-spectrum antimicrobial therapy within the first hour of recognition (4). A recent prospective study comparing an aggressive antimicrobial treatment protocol (i.e., initiation of empiric antibiotics for the suspicion of infection after blood cultures were drawn) to a conservative antimicrobial protocol (i.e., antimicrobial initiation was withheld until objective culture-proven data were obtained) among a critically ill surgical population observed that the aggressive treatment protocol resulted in reduced appropriateness of initial antimicrobial therapy, prolonged duration of antimicrobial therapy, and a significant reduction in survival compared with the conservative treatment protocol (17).
One-thousand four-hundred ninety-eight of the 2,085 inappropriately treated infections were ultimately changed to appropriate antimicrobial therapy once culture results were obtained. Patient survival would be expected to improve with appropriate therapy, and this may have resulted in the observed findings of our study. In addition, some studies have shown that early adequate antibiotic therapy may only improve survival in patients with minimal illness severity, whereas, in other patients, the baseline severity is so great that no benefit is appreciated (9). Among the appropriately treated infections in our study, an increased prevalence of younger trauma patients with greater APACHE II scores were observed compared with inappropriately treated infections. Some studies have shown that APACHE II scores may underestimate the true mortality risk in many trauma patients secondary to an increased prevalence of younger patients who lack chronic health problems commonly observed in the older population (31, 32).
Strengths and limitations
Our study is strengthened by its large sample size and multivariable analysis; however, it may be limited by its retrospective design (e.g., selection bias and confounding). This was a single-center study, and thus, external validity may be limited in generalizing results to other areas as the demographics and comorbidities of our patient population may differ. Cause of death was not captured by our database. Thus, it is possible that mortality was caused by factors unrelated to sepsis and/or antibiotic inappropriateness. Furthermore, only 12.4% of inpatient deaths at our institution, during the study period, received postmortem autopsies. Previous studies have observed a wide discrepancy between clinical and postmortem findings attributable to the cause of death, especially within surgical ICU, trauma, and transplant patients (33–35). Although time from sepsis recognition to antibiotics (appropriate and inappropriate) was captured by our database, time to ultimate appropriate antibiotic treatment was not. Previous studies have observed a significant association between delay in treatment with appropriate antibiotics and mortality (15). In addition, although our database captured APACHE II scores at time of each infectious episode, it is not explicitly clear at what point during the episode this occurred. Previous studies recommend that the optimal time to record the severity of illness be just before the true onset of bacteremia (i.e., 48 h before collection of the initial blood sample obtained for culture) (15).
In contrast to our hypothesis, our study observed no difference in mortality between infections initially treated with either appropriate or inappropriate antimicrobial therapy. The difference in our data and those previously reported might lie in the fact that all of our patients were treated on a surgical service, where a large percentage of infections are treated primarily by mechanical source control procedures rather than antimicrobials alone. Perhaps “appropriateness” in empiric antimicrobial therapy is less critical in surgical patients compared with timing of antimicrobial therapy and/or source control. Until biomarker identification, metabolomic profiling, and/or rapid microbiologic microarrays become mainstream, an awareness of hospital microbial flora/fauna and patient acuity, combined with rapid intervention (i.e., source control), close attention to culture results with subsequent deescalation, and infectious disease involvement should remain paramount regarding standard of care in this surgical patient population.
1. Gaieski DF, Edwards JM, Kallan MJ, Carr BG: Benchmarking the incidence and mortality of severe sepsis in the united states. Crit Care Med
41 (5): 1167–1174, 2013.
2. Bone RC, Balk RA, Cerra FB, Dellinger RP, Fein AM, Knaus WA, Schein RM, Sibbald WJ: Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine. Chest
101 (6): 1644–1655, 1992.
3. Lagu T, Rothberg MB, Shieh MS, Pekow PS, Steingrub JS, Lindenauer PK: What is the best method for estimating the burden of severe sepsis in the United States? J Crit Care
27 (4): 414 e1–e9, 2012.
4. Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach H, Opal SM, Sevransky JE, Sprung CL, Douglas IS, Jaeschke R, et al.: Surviving sepsis campaign: international guidelines for management of severe sepsis and septic shock: 2012. Crit Care Med
41 (2): 580–637, 2013.
5. Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J, Pinsky MR: Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit Care Med
29 (7): 1303–1310, 2001.
6. Aarts MA, Granton J, Cook DJ, Bohnen JM, Marshall JC: Empiric antimicrobial therapy in critical illness: results of a Surgical Infection Society survey. Surg Infect
8 (3): 329–336, 2007.
7. Kuti EL, Patel AA, Coleman CI: Impact of inappropriate antibiotic therapy on mortality in patients with ventilator-associated pneumonia and bloodstream infection: a meta-analysis. J Crit Care
23 (1): 91–100, 2008.
8. Ramphal R: Importance of adequate initial antimicrobial therapy. Chemotherapy
51 (4): 171–176, 2005.
9. Clec’h C, Timsit JF, De Lassence A, Azoulay E, Alberti C, Garrouste-Orgeas M, Mourvilier B, Troche G, Tafflet M, Tuil O, et al.: Efficacy of adequate early antibiotic therapy in ventilator-associated pneumonia: influence of disease severity. Intensive Care Med
30 (7): 1327–1333, 2004.
10. Tellado JM, Sen SS, Caloto MT, Kumar RN, Nocea G: Consequences of inappropriate initial empiric parenteral antibiotic therapy among patients with community-acquired intra-abdominal infections in Spain. Scand J Infect Dis
39 (11-12): 947–955, 2007.
11. Kollef MH: Broad-spectrum antimicrobials and the treatment of serious bacterial infections: getting it right up front. Clin Infect Dis
47 (Suppl 1): S3–S13, 2008.
12. Kollef MH, Sherman G, Ward S, Fraser VJ: Inadequate antimicrobial treatment of infections: a risk factor for hospital mortality among critically ill patients. Chest
115 (2): 462–474, 1999.
13. Krobot K, Yin D, Zhang Q, Sen S, Altendorf-Hofmann A, Scheele J, Sendt W: Effect of inappropriate initial empiric antibiotic therapy on outcome of patients with community-acquired intra-abdominal infections requiring surgery. Eur J Clin Microbiol Infect Dis
23 (9): 682–687, 2004.
14. Sturkenboom MC, Goettsch WG, Picelli G, in’t Veld B, Yin DD, de Jong RB, Go PM, Herings RM: Inappropriate initial treatment of secondary intra-abdominal infections leads to increased risk of clinical failure and costs. Br J Clin Pharmacol
60 (4): 438–443, 2005.
15. McGregor JC, Rich SE, Harris AD, Perencevich EN, Osih R, Lodise TP Jr, Miller RR, Furuno JP: A systematic review of the methods used to assess the association between appropriate antibiotic therapy and mortality in bacteremic patients. Clin Infect Dis
45 (3): 329–337, 2007.
16. Horan TC, Andrus M, Dudeck MA: CDC/NHSN surveillance definition of health care–associated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control
36 (5): 309–332, 2008.
17. Hranjec T, Rosenberger LH, Swenson B, Metzger R, Flohr TR, Politano AD, Riccio LM, Popovsky KA, Sawyer RG: Aggressive versus conservative initiation of antimicrobial treatment in critically ill surgical patients with suspected intensive-care-unit-acquired infection: a quasi-experimental, before and after observational cohort study. Lancet Infect Dis
12 (10): 774–780, 2012.
18. Gilbert DN, Moellering RC, Eliopoulos GM, Chambers HF, Saag MS (eds.): The Sanford Guide to Antimicrobial Therapy
. 41st ed. Sperryville, VA: Antimicrobial Therapy, Inc., p 220, 2011.
19. Petersen MR, Deddens JA: A comparison of two methods for estimating prevalence ratios. BMC Med Res Method
8: 9, 2008.
20. Valles J, Leon C, Alvarez-Lerma F: Nosocomial bacteremia in critically ill patients: a multicenter study evaluating epidemiology and prognosis. Spanish Collaborative Group for Infections in Intensive Care Units of Sociedad Espanola de Medicina Intensiva y Unidades Coronarias (SEMIUC). Clin Infect Dis
24 (3): 387–395, 1997.
21. Rello J, Ricart M, Mirelis B, Quintana E, Gurgui M, Net A, Prats G: Nosocomial bacteremia in a medical-surgical intensive care unit: epidemiologic characteristics and factors influencing mortality in 111 episodes. Intensive Care Med
20 (2): 94–98, 1994.
22. Leibovici L, Shraga I, Drucker M, Konigsberger H, Samra Z, Pitlik SD: The benefit of appropriate empirical antibiotic treatment in patients with bloodstream infection. J Intern Med
244 (5): 379–386, 1998.
23. Turnbull IR, Javadi P, Buchman TG, Hotchkiss RS, Karl IE, Coopersmith CM: Antibiotics improve survival in sepsis independent of injury severity but do not change mortality in mice with markedly elevated interleukin 6 levels. Shock
21 (2): 121–125, 2004.
24. Deresinski S: Principles of antibiotic therapy in severe infections: optimizing the therapeutic approach by use of laboratory and clinical data. Clin Infect Dis
45 (Suppl 3): S177–S183, 2007.
25. Marino PL: In: Brown B, Dernoski N, Lazar T (eds.): The ICU Book
. 3rd ed. Philadelphia, PA: Lippincott Williams and Wilkins, pp 769–782, 2007.
26. Montagna MT, Caggiano G, Lovero G, De Giglio O, Coretti C, Cuna T, Iatta R, Giglio M, Dalfino L, Bruno F, et al.: Epidemiology of invasive fungal infections in the intensive care unit: results of a multicenter Italian survey (Aurora Project). Infection
41 (3): 645–653, 2013.
27. Leroy O, Gangneux JP, Montravers P, Mira JP, Gouin F, Sollet JP, Carlet J, Reynes J, Rosenheim M, Regnier B, et al.: Epidemiology, management, and risk factors for death of invasive Candida
infections in critical care: a multicenter, prospective, observational study in France (2005–2006). Crit Care Med
37 (5): 1612–1618, 2009.
28. Tabah A, Koulenti D, Laupland K, Misset B, Valles J, Bruzzi de Carvalho F, Paiva JA, Cakar N, Ma X, Eggimann P, et al.: Characteristics and determinants of outcome of hospital-acquired bloodstream infections in intensive care units: the Eurobact International Cohort Study. Intensive Care Med
38 (12): 1930–1945, 2012.
29. Figueiredo Costa S: Impact of antimicrobial resistance on the treatment and outcome of patients with sepsis. Shock
30 (Suppl 1): 23–29, 2008.
30. Samraj RS, Zingarelli B, Wong HR: Role of biomarkers in sepsis care. Shock
40 (5): 358–365, 2013.
31. Vassar MJ, Wilkerson CL, Duran PJ, Perry CA, Holcroft JW: Comparison of APACHE II, TRISS, and a proposed 24-hour ICU point system for prediction of outcome in ICU trauma patients. J Trauma
32 (4): 490–499, 1992.
32. McNelis J, Marini C, Kalimi R, Jurkiewicz A, Ritter G, Nathan I: A comparison of predictive outcomes of APACHE II and SAPS II in a surgical intensive care unit. Am J Med Qual
16 (5): 161–165, 2001.
33. Shojania KG, Burton EC, McDonald KM, Goldman L: Changes in rates of autopsy-detected diagnostic errors over time: a systematic review. JAMA
289 (21): 2849–2856, 2003.
34. Ong AW, Cohn SM, Cohn KA, Jaramillo DH, Parbhu R, McKenney MG, Barquist ES, Bell MD: Unexpected findings in trauma patients dying in the intensive care unit: results of 153 consecutive autopsies. J Am Coll Surg
194 (4): 401–406, 2002.
35. Mort TC, Yeston NS: The relationship of premortem diagnoses and postmortem findings in a surgical intensive care unit. Crit Care Med
27 (2): 299–303, 1999.