Reply to Llor and Butler.

TO THE EDITOR—We thank Llor and Butler [1] for their thoughtful comments regarding point-of-care (POC) testing in the primary care setting. The Infectious Diseases Society of America (IDSA) policy paper [2] was designed to include a wide spectrum of clinical practice settings, including outpatient, emergency department , inpatient, and intensive care units. As a practical matter, we arrived at 1 hour as a reasonable turnaround time for the test result when considering many clinical settings. We agree with the authors that to impact clinical care in the outpatient or primary care setting, the availability of a test result within 5–15 minutes would be ideal, and is likely needed for the test to be widely adopted. Depending on the patient flow in an office setting, a test with up to a 30-minute turnaround time may also be useful. As indicated in both the letter and our paper, POC tests will be most beneficial if they are accurate, simple to use, and, ideally, inexpensive. Whereas we emphasize making an eti-ologic diagnosis in some clinical settings and for certain specific pathogens, we also highlight the value of a test that can rule in or rule out infection, or one that can quickly distinguish between bacterial and viral pathogens, particularly in the outpatient setting. Testing would not necessarily need to identify the causative pathogen, and it is likely that host bio-markers would play an important role. Although we do not specifically mention C-reactive protein testing, we agree it is an example of a biomarker test that may meet the clinical need in some situations. Realistically, tests that are used to rule out a potentially serious bacterial infection will need to have a very high negative predictive value (we suggest ≥98%) to change clinical practice and provide healthcare providers with the confidence to withhold antibiotics. Many patients with infections are initially seen by a primary care physician, and the availability of tests that can impact clinical decision making in the office setting is clearly important. We concur with the authors that outcomes research is essential to show the value of new diagnos-tics in a variety of clinical settings. Indeed, one of IDSA's diagnostics policy priorities is to encourage the funding and conduct of outcomes research, including clinical 1488 • CID 2014:58 (15 May) • CORRESPONDENCE