590. Mycobacteriology Diagnostic Costs in an Inner-City Public Hospital: Are We Ready to Adopt the Xpert® MTB/RIF in Lieu of the Acid-Fast Bacilli (AFB) Smear?
Session: Poster Abstract Session: TB: Mycobacteria Diagnostic Testing
Thursday, October 8, 2015
Room: Poster Hall
Background: Grady Memorial Hospital (GMH) in Atlanta, GA is a 650-bed urban safety-net hospital in a high tuberculosis (TB) incidence area. Our current mycobacterial diagnostic practice involves AFB smear and culture (cx) for all samples.  Xpert® MTB/RIF is automatically performed on smear positive samples (“selective Xpert” algorithm”). We sought to compare our current algorithm with (a) “universal Xpert, no smear” approach in which all samples are submitted for culture and only 1 sample per patient is tested with Xpert® MTB/RIF (b) “universal Xpert with smear” algorithm in which, all samples are processed for culture and smear and only 1 sample per patient is tested with Xpert® MTB/RIF.

Methods: Lab records for patients with respiratory specimens submitted for smear and cx from October 1 to December 31, 2014 were reviewed.  Estimated costs (in US$) of smear, cx, and Xpert were based on published cost analyses (table1). Patients diagnosed by Xpert were defined as “early TB diagnosis”; patients diagnosed by culture only were defined as “delayed TB diagnosis”. Non-tuberculous mycobacteria (NTM) disease was defined per national guidelines.

Results: 319 patients submitted 650 specimens (2.03 per patient). 10 patients were diagnosed with TB; 7 had a positive smear. 3 had NTM disease. Both universal Xpert approaches, with or without smear, would avert 2 delayed TB diagnoses; a 66.6% reduction compared to the “selective Xpert” algorithm (figure 1). A low cost Xpert could make algorithms that do not involve smears attractive from a laboratory cost perspective (table2).

Conclusion: The addition of universal Xpert to our current algorithm would reduce delays in TB diagnosis, which may have significant benefit to both individual and public health. There are negative individual health consequences with the “universal Xpert, no smear” approach.

Table1. Laboratory cost ranges


Base cost

Low cost

High cost













Table2. Accrued cost per algorithm


Base cost

Low cost

High cost

Selective Xpert




Universal Xpert, no smear




Universal Xpert and smear




Marcos C. Schechter, MD1, Gretchen Snoeyenbos, MD2, Yun F. Wang, MD, PhD3,4, David P. Holland, MD, MHS1 and Susan M. Ray, MD1, (1)Department of Medicine, Division of Infectious Diseases, Emory University School of Medicine, Atlanta, GA, (2)Department of Medicine, Emory University School of Medicine, Atlanta, GA, (3)Department of Pathology and Laboratory Medicine, Emory University School of Medicine, Atlanta, GA, (4)Grady Memorial Hospital, Atlanta, GA


M. C. Schechter, None

G. Snoeyenbos, None

Y. F. Wang, None

D. P. Holland, None

S. M. Ray, None

Findings in the abstracts are embargoed until 12:01 a.m. PDT, Wednesday Oct. 7th with the exception of research findings presented at the IDWeek press conferences.