Update in Hospital Medicine

  1. Preetha Basaviah, MD; and
  2. Shaun Frost, MD
  1. From Stanford University, Stanford, and Cogent Healthcare, Irvine, California.

    2006 Series: Update Sessions from ACP's 2006 Annual Session

    We selected the articles for this Update in Hospital Medicine after surveying the literature and consulting with opinion leaders from across the United States. Criteria for selection were quality of evidence and importance to hospital medicine practitioners and hospital administrators. The articles are divided into those that are likely to affect clinical practice and are relevant to patient safety and quality improvement. The American College of Physicians presented this Update in Hospital Medicine in cooperation with the Society for Hospital Medicine. Changes to clinical practice emerging from these articles are shown in the Table .

    View this table:
    Table. Changes to Clinical Practice Emerging from Articles Important to Hospitalists in 2005*

    Clinical Medicine

    Paired Quantitative Blood Culture Was Best for Identifying Intravascular Device–Associated Bloodstream Infection

    Bloodstream infections from intravascular devices (IVDs), such as percutaneous catheters and subcutaneous ports, are a clinically significant source of illness and death. Of the more than 200 000 cases of IVD-related bloodstream infection that occur annually in the United States, 12% to 25% result in death (4). Detecting IVD-related bloodstream infection without removing the device is difficult because the most sensitive findings of bloodstream infection, such as fever, are not specific, and the most specific findings, such as pus around a central line site, are not sensitive. In at least 2 studies, 80% of catheters removed because of possible infection were found to be uninfected (5, 6).

    Safdar and colleagues' meta-analysis sought to determine the most accurate methods for identifying IVD-related bloodstream infection without IVD removal. They reviewed the literature and identified 51 studies of 8 frequently used diagnostic tests for IVD-related bloodstream infection that were published between 1966 and 31 July 2004. The studies included comparisons with a reference standard and provided data for calculating sensitivity, specificity, and likelihood ratios.

    The 8 diagnostic tests were catheter segment culture (qualitative, semiquantitative, and quantitative), IVD-drawn blood culture (qualitative, quantitative, and quantitative paired peripheral and IVD-drawn), acridine …

    Responses to this article

    « Previous | Next Article »Table of Contents