Faculty Bibliography

BACKGROUND:Conflicting evidence exists regarding probiotics and the incidence of Clostridioides difficile infection (CDI). This study evaluates whether probiotics are efficacious for CDI prophylaxis in patients receiving antibiotics. METHODS:A retrospective cohort analysis of patients admitted to NYU Winthrop Hospital who received at least 1 dose of antibiotics considered high risk of inducing CDI. Patients were grouped according to probiotic use; association between probiotic use and incident CDI was examined. A model for incident CDI adjusting for known CDI risk factors was estimated. RESULTS:Of 3,267 patients, 4.6% had CDI within 12 weeks of antibiotics initiation. A total of 5.1% received probiotics within 24 hours of initiation, and 6.6% initiated probiotics during the 12-week follow-up. Of those taking probiotics within 24 hours of antibiotics, 9.6% had CDI, and of those not taking probiotics 4.2% had CDI (relative risk, 2.3; 95% confidence interval, 1.4, 3.7). In time-dependent Cox models accounting for probiotic initiation and adjusting for potential confounders, a positive association between probiotics and CDI remained significant (hazard ratio, 2.7; P < .001). DISCUSSION/CONCLUSIONS:Patients who received antibiotics with concurrent probiotics were more likely to have an incident of CDI compared with those who did not receive probiotics. Additional risk factors were histamine 2 receptor antagonists, proton pump inhibitors, and administration of multiple antibiotics simultaneously. CONCLUSIONS:The present study, because of its large population and inclusion of multiple variables playing a role in CDI, serves as a valuable resource when considering efficacy of probiotics as CDI prophylaxis.

Traditionally, there have been uniform antibiotic dosing guidelines for prophylaxis for clean-clean-contaminated surgery in both non-obese and obese adults. All other factors predisposing to surgical site infections (SSIs) being equal, over time, the preferred drug is cefazolin. The usual dose, given immediately pre-procedure, has been 1 g intravenously (IV) in non-penicillin-allergic patients, which has been highly effective, Recently, it has become common practice to use high-dose cefazolin; i.e., 3 g IV, in obese patients. This article reviews the literature on high-dose cefazolin prophylactic regimens in the obese from a pharmacokinetic (PK) point of view. There are no comparative studies to support this approach, which is based largely on the theory "more must be better." Weight-based dosing of cefazolin in the obese is flawed, because it does not take into account PK factors, which are critical in the obese. Cefazolin is a water-soluble (hydrophilic) antibiotic that does not penetrate adipose tissue regardless of IV dose. Importantly, adipose tissue is not a valid target tissue in clean-clean-contaminated SSI prophylaxis, as it does not become infected. Higher doses result in proportionately higher serum/non-adipose tissue concentrations, but adipose tissue concentrations are unaffected. Cefazolin displays time-dependent killing kinetics so that as long as serum/tissue concentrations are above the minimum inhibitory concentration (MIC) of SSI pathogens, there is no enhanced killing with higher concentrations relative to concentration-dependent antibiotics. Taking into account PK principles, a cefazolin 1 g IV bolus results in peak serum concentrations of ∼185 mcg/mL, provides at least six hours of intra-operative protection, aside from any post-antibiotic effects, and eliminates any rationale for intra-operative re-dosing for procedures lasting six hours or less. Some have argued that a cefazolin 3 g IV dose in the obese does not matter, as more must necessarily be better. However, from an antibiotic stewardship program (ASP) perspective, unneeded antibiotics are unnecessary. Moreover, the costs of cefazolin 1 g (IV push) at $0.75 versus 2 g (IV piggyback) at $ 6.83 can be significant in large centers using cefazolin prophylaxis for cardiothoracic, orthopedic, obstetric/gynecology, and bariatric surgery. Excessive antibiotics also expose the patient to potential adverse effects; i.e., Clostridium difficile. There is no dose-dependent or duration of exposure effect on resistance with one or two pre-operative or intra-operative doses. Well-done PK-based studies in obese patients clearly demonstrate the lack of benefit of using a 3-g dose or intra-operative re-dosing and show no incremental increase in adipose tissue concentrations with high doses. From an ASP point of view, antibiotic dosing recommendations should be reviewed and revised on the basis of PK principles that indicate that weight-based dosing has no basis for pre-operative prophylaxis in obese patients.

Effective antimicrobial therapy depends on several factors including degree of activity against the pathogen, antibiotic resistance, and when relevant, optimal tissue penetration factors. Central nervous system (CNS) infections illustrate these points well. The pharmacokinetic (PK) parameters important in antibiotic blood cerebrospinal fluid barrier (BCB) penetration that is important in meningitis are different and do not predict blood brain barrier (BBB) penetration. Recently, we had a case of Mycoplasma pneumoniae encephalitis (MPE) which prompted a review of the antibiotic PK determinants of BBB penetration which differ markedly from those of BCB penetration important in encephalitis. Using MPE as an illustrative example, this article reviews host and drug factors of therapeutic importance in optimally treating MPE.

Tetracycline antimicrobials entered into clinical usage in the late 1940s. Permanent dental staining from tetracyclines was first appreciated in 1956, eventually leading to avoidance of this class of antibiotics whenever possible in young children and pregnant or breastfeeding women. Doxycycline, introduced in 1967, binds calcium less avidly than prior tetracyclines and is regarded by some authorities as safe to prescribe for pregnant women and young children. Review of the available data, however, suggests that this interpretation may be incorrect or at least premature. In conclusion, until more definitive data are developed, doxycycline should continue to be only selectively prescribed for young children and pregnant or breastfeeding women for whom alternative, safer antibiotics are not available, and courses of treatment should be of as short a duration as possible.

Sir William Osler is revered as a consummate clinician, extraordinary teacher, and caring mentor. Importantly, he advocated and practiced humanistic medicine. Osler taught that in the golden age of Greece, medicine had a relationship to the humanities. Hopefully, exposing medical residents and practitioners to the humanities will foster empathy in patient care. Clinical excellence and compassionate care should be what medical care is all about. Osler believed that reading in the humanities contributed to the education of the heart. Osler pointed the way, as always leading by personal example. He understood that clinical practice leaves precious little time for personal development. Oslers practical solution for himself and for all physicians was selected reading in the classics. His bedside library consisted of 10 books (Oslers 10) which he particularly valued as a way to understand the human condition. Reading in the humanities enhances appreciation of human suffering. Since Osler, others, including myself, have added or substituted other classics or humanities books besides Oslers original 10, reflecting the physicians own interests and preferences. My own bedside library contains five books from Oslers 10 as well as some personal favorites. There are no must reads. Each physician should select his or her own bedside library based on their personal interests or relevance. Osler led the way in suggesting that a practical solution for busy practicing physicians to enhance empathy for more compassionate patient care is by reading in the humanities.

Clinical correlation is essential in assessing the relevance of the patient's history and physical findings in making a clinical presumptive diagnosis. False diagnostic associations may result in misdiagnosis. We present a case of an elderly female with HIV on HAART who presented with shortness of breath assumed to have Pneumocystis (carinii) jiroveci pneumonia (PCP) even though she had a clinical diagnosis of influenza B. She was thought to have PCP only because she had HIV. Tests for PCP were negative including BAL staining. Influenza B present in her respiratory secretions by PCR and was also cultured from BAL fluid. Diagnostic associations are helpful in suggesting diagnostic possibilities but must be supported by clinical correlation of characteristic clinical features.

Legionnaire's disease (LD) is a non-zoonotic atypical community acquired pneumonia (CAP) with several characteristic extra-pulmonary findings. Pending diagnostic test results, selected characteristic findings when considered together are the basis of clinical syndromic diagnosis and the basis of empiric antimicrobial therapy. Of the extra-pulmonary manifestation of LD, neurologic findings are among the most common, e.g., headache, mental confusion. In LD, encephalitis is rare as are myoclonus and seizures. This is a most interesting case of LD that presented with encephalitis, myoclonus and seizures. Pulmonary infiltrates developed early after admission. LD was suspected on the basis of otherwise unexplained characteristic findings, e.g., hypophosphatemia, elevated serum transaminases, microscopic hematuria, elevated ferritin, and empiric doxycycline therapy was started. The diagnosis of LD was further supported by prominent and persistent myoclonus and seizures, rare but characteristic neurologic findings in LD. On week 12 of hospitalization, he finally seroconverted with negative urinary antigen tests indicating his LD was due to a non-L. pneumophilia (serotype 01) strain. On doxycycline, he made a slow but complete recovery. We believe this is the first reported case of LD presenting with encephalitis, myoclonus, and seizures successfully treated with doxycycline.