In the Literature Free

Every Day Of Antipseudomonal β-Lactam Antibiotic Administration Is Associated With A Risk Of Selection Of Resistant Pathogens

Teshome BF, Vouri SM, Hampton N, Kollef MH, Micek ST. Duration of Exposure to Antipseudomonal β-Lactam Antibiotics in the Critically Ill and Development of New Resistance. Pharmacotherapy. 2019 Mar;39(3):261–270. doi: 10.1002/phar.2201. Epub 2019 Jan 7. PubMed PMID: 30506852.

Teshome and colleagues in St. Louis, Missouri, examined the relationship of the duration of administration of broad spectrum ß-lactam antibiotics (cefepime, meropenem, and piperacillin-tazobactam) to the selection of bacterial isolates resistant to 1 or more of these in adults with severe sepsis or septic shock between 2010 and 2015. In this single center retrospective cohort study, resistance was evaluated relative to the specific antibiotic and not to individual pathogens. The appearance of a previously undetected (in the 180 days prior to antibiotic initiation) pathogen that was resistant to 1 or more antipseudomonal ß-lactam meant that selection of a resistant organism was considered to have occurred. Although not evaluated, the emergence of resistance could have resulted from genetic mutation, the acquisition of new genetic material, or, as seems likely in most cases, simple selection of preexisting resistant organisms that became detectable under the selective pressure of antibiotic exposure.

The study population consisted of 7118 patients who had received at least 1 dose of an antipseudomonal ß-lactam, most often cefepime (5274), followed by meropenem (2463) and piperacillin-tazobactam (2463). New resistance was detected in 444 (6.2%) patients with a median time to appearance of 17 days from the first dose of an antipseudomonal ß-lactam. The rate of resistance detection was 0.16 per 100 patient-days; there was a 4% increased risk of new resistance for each day of exposure to an antipseudomonal ß-lactam. In examining each administered antibiotic, the increased risk per day of exposure was 8% for cefepime, 2% for meropenem, and 8% for piperacillin-tazobactam. In illustrating the significance of these rates, the authors point out that for cefepime, for example, a 10-day course of administration is associated with a 24% increased risk of selection of resistance to an antipseudomonal ß-lactam when compared to a 7-day course.

Although a number of issues could be raised regarding this study, the results are consistent with other evaluations (and with common sense) indicating that the duration of administration is a critical time-dependent variable in the of selection of antibiotic resistance organisms. It supports efforts, including a required implementation of antibiotic time-outs, aimed at shortening the duration of unnecessary antibiotic exposure, with the aim of preventing unnecessary adverse effects of the antibiotics, including the selection of resistant pathogens.

Not All Sepsis Deaths Are Preventable

Rhee C, Jones TM, Hamad Y, Pande A, Varon J, O’Brien C, et al; Centers for Disease Control and Prevention (CDC) Prevention Epicenters Program. Prevalence, Underlying Causes, and Preventability of Sepsis-associated Mortality in US Acute Care Hospitals. JAMA Netw Open. 2019 Feb 1;2(2):e187571. doi: 10.1001/jamanetworkopen.2018.7571. PubMed PMID: 30768188.

Sepsis, which is reported to contribute to more than a quarter million deaths each year in the United States, is present in as many as one-half of hospitalized patients who die. Among the important considerations in considering the relationship between sepsis and mortality is the extent to which sepsis actually contributes to death in these patients and, importantly, what proportion of sepsis-associated deaths are preventable. Rhee and colleagues have addressed the second of these questions.

In this retrospective cohort study, the investigators examined the medical records of 568 randomly selected adults with sepsis (Sepsis-3 criteria) at 3 academic and 3 affiliated community hospitals over 24 months ending 31 December 2015, who died during the hospitalization or were discharged to hospice care and were not readmitted. The latter group accounted for approximately one-half of the total, 289/568 (50.9%), and approximately one-third of patients in hospice died within 7 days of admission.

Sepsis was present on admission in 221 (73.7%) patients and was of hospital-onset in the remaining 79 (26.3%.). As judged by hospice criteria, end-stage comorbidities (eg, progressive solid or hematologic cancer, advanced dementia) were present in 121 (40.3%) of the patients who died. Sepsis was judged to be the immediate cause of death in 198 (34.9%) patients, a proportion much greater than, for example, progressive cancer (16.2%) and heart failure (6.9%) in this regard. Approximately one-half of patients in whom sepsis was the immediate cause of death had pneumonia, whereas one-fifth had intra-abdominal infection.

Preventability of death was assessed by review performed by 2 clinicians and included use of a Likert scale. In total, 264 (88.0%) fatalities were considered unpreventable leaving only 36 (12.0%) who were potentially preventable. Of these 36, 25 (8.3%) were possibly preventable, and only 11 (3.7%) were definitely or moderately likely preventable. Suboptimal care was judged to have taken place in 68 (22.7%) of the total of 300 patients, most frequently because of delay in initiation of antibiotic therapy in 33 (48.5%), a delay in source control in 19 (27.9%), and inappropriate antibiotic selection in relation to the subsequent antibiotic susceptibility report in 16 (23.5%). Forty-two major errors occurred in the 36 patients with potentially preventable death; 10 of these were the result of initiation of inappropriate antibiotics.

This study confirms that most sepsis-associated deaths in adults occur in medically complex patients with potentially fatal comorbidities. The most frequently identified shortcomings in care were a delay in appropriate antibiotic initiation and in source control. It also demonstrates that the proportion of deaths that are possibly or definitely preventable was only 12%, and only 11 (3.7%) deaths were moderately or definitely preventable. These observations demonstrate that, at least in highly effective medical care facilities, we may be approaching the limits of what can be accomplished with current therapeutic approaches to sepsis.

Editor’s Note (contributed by Ferric C. Fang, MD – Deputy Editor): The small proportion of preventable sepsis-related deaths observed in this study is related to the small proportion who received IEAT (inappropriate empirical antibiotic treatment, ie, mismatched to organism susceptibility), which was observed in just 16 subjects comprising 5.3% of sepsis-associated deaths. This is considerably lower than the 32% of patients observed to receive IEAT for severe bacterial infections in a meta-analysis of nearly 75 000 patients in 191 studies (Carrara, et al. Int J Antimicrob Agents 51:548–53, 2018). As IEAT is associated with the prevalence of multidrug-resistant Gram-negative organisms and mortality rates, one might have expected the proportion to be even higher in this cohort of patients with sepsis-associated death. However, as microbiological data were not provided, the reasons for the unexpectedly low proportion of IEAT cannot be ascertained. The conclusion that few sepsis-associated deaths are preventable may not be applicable to other settings with higher rates of MDROs and IEAT.

Case Vignette: Parvovirus B19 And Acute Hepatic Failure

Islek A, Keskin H, Agın M, Aksungur N, Korkut E, Ozturk G. Parvovirus B19 Infection as a Rare Cause of Fulminant Liver Failure: A Case Report. Transplant Proc. 2019 May;51(4):1169–1171. doi: 10.1016/j.transproceed.2019.01.084. Epub 2019 Feb 7. PubMed PMID: 31101193.

Islek and colleagues describe 3 previously healthy children with fulminant liver failure as a consequence of acute infection with parvovirus B19. A 5-year-old girl developed a febrile illness with skin rash, and 6 days later, blood tests demonstrated evidence of severe acute hepatocellular injury. Parvovirus B19 immunoglobulin M (IgM) antibody was positive as was polymerase chain reaction (PCR). She developed progressive hepatic failure and underwent successful liver transplantation. Testing after 3 months failed to detect circulating parvovirus DNA.

A 14-year-old boy was admitted with fatigue and jaundice 2 weeks after an upper respiratory tract infection. Serum ALT and AST were 1810 U/L and 1680 U/L, respectively, and bilirubin was 13 mg/dL. Both parvovirus B19 IgM antibody and PCR were positive. He developed progressive hepatic failure and then underwent successful liver transplantation. Parvovirus PCR was negative 10 weeks after transplantation. A 16-year-old boy, the brother of the 14-year-old, also presented with acute hepatic failure after an upper respiratory tract infection and had the same evidence for acute parvovirus infection. He, however, spontaneously recovered.

Of note is that there is no mention of arthralgias in these patients, 2 of whom apparently had no skin rash—findings consistent with the lack of immunoglobulin G antibody directed against the virus on presentation. Somewhat surprisingly, none of the 3 were anemic (hemoglobin levels were 12.6–14.4 g/dL). All had negative studies for other causes of acute hepatic failure. There is no mention in the case reports of treatment with intravenous immunoglobulin, although all 3 patients underwent plasma exchange. All managed to clear the virus despite administration of immunosuppressive therapy to prevent rejection of the transplanted liver in the first 2 cases.

The authors indicate that the number of cases of acute hepatitis due to parvovirus B19 reported in the English literature is quite small. The occurrence of this syndrome in 2 brothers raises the issue of an inherited genetic abnormality predisposing to severe infection with parvovirus B19. Testing for parvovirus B19 should be part of the search for an etiology of fulminant hepatic failure.