Lance Peterson

Utilization and Performance of a Laboratory Developed Nucleic Acid Amplification Test for the Diagnosis of Pulmonary and Extrapulmonary Tuberculosis in a Low Prevalence Area: A 14 Year Study.

Abstract Background Tuberculosis (TB) is a significant global health problem. Nucleic acid amplification tests (NAATs) are valuable in reducing delays to initiation of therapy and infection control protocols. A retrospective study was performed to assess the utilization and performance of a laboratory developed Mycobacterium tuberculosis complex (MTBC) PCR assay (TBPCR) for diagnosis of pulmonary (PTB) and extrapulmonary (EPTB) tuberculosis. Methods Study site was a 4 hospital system in suburban Chicago. All culture confirmed TB specimens with complete laboratory data from January 2002 to December 2016 were included. Patient records were accessed using an electronic data warehouse, following approval from Institutional Review Board. Standard microbiology procedures were followed for smear and culture of MTBC. A lab-developed real time PCR targeting a 123 bp region of the IS6110 insertion sequence of MTBC was performed on smear positive specimens or if ordered by physician. Clinical and laboratory data was compared with TBPCR results for all culture confirmed cases. Results There were 151 culture positive patients and 2186 TBPCR performed. Median age of patients at diagnosis was 49 years (IQR 33–66), 74 (49%) were female and 14 were on immunosupressive therapy. The mean number of samples tested per patient was 2. Of culture positive specimens, 59% were from a respiratory source and 3 were MDR; ordering of TBPCR was higher in specimens from PTB source (58.4%) as compared with EPTB source (37%). Combined sensitivity of the TBPCR on all specimen types was 86.6% (95% CI 76.3–93.1); 90.3% for PTB specimens alone (95% CI 78.2–96.4). Specificity was 100% (95% CI 99.5–100), PPV 100% (95% CI 90.5–100%) and NPV 99.5% (95% CI 98.8–99.8%), and were similar for all specimen types. Sensitivity of TBPCR was 97% in smear positive and 79% in smear negative PTB specimens. The median time to culture positivity was 7 days longer in specimens that were TBPCR negative compared with those that were positive (P = 0.14, NS), however, TBPCR shortened time to diagnosis by 13 days. Conclusion We found TBPCR to be underutilized in both PTB and EPTB although it was found to be a rapid and reliable method for early diagnosis. Education regarding utility of NAATs could be useful in low burden areas where paucibacillary disease is more common, especially in EPTB. Disclosures All authors: No reported disclosures.

1035New Urine Reporting Criteria to Accurately Report Nosocomial Clinical Urinary Tract Infection

Background. Reducing unnecessary antimicrobial therapy is critical to patient safety. We previously conducted an analysis to establish a colony count threshold predicting clinically significant UTI developing in hospitalized patients (AJCP 2012;137:778-84). Patients with urine culture colony counts >10 CFU/ml were 74 times more likely to have a clinically significant UTI than patients with colony counts <10 CFU/ml. With the approval of the Departments of Urology, Infectious Disease, Quality, and Infection Control we modified our urine culture laboratory reporting criteria for voided and Foley catheter samples from hospitalized patients with a length of stay of >2 days. For these patients, a positive test consists of 1-2 organisms at >10CFU/ml. Any other colony count or mixture of bacteria is reported as “Negative for Nosocomial UTI” (NNUTI). We hypothesize that this new reporting scheme would accurately report the absence of a UTI in >95% of samples. The first 5 months of the new reporting approach was validated with chart review. Methods. Inpatient urine cultures were assessed to determine if 1) a patient had been in the hospital >2 days when the culture was taken and 2) the urine was a voided or Foley sample. The culture report was assessed with chart review by a single Infectious Disease Physician to determine if the patient had signs and symptoms of a UTI when NNUTI was the result. Parameters to determine UTI included fever >100.4°F, frequency, dysuria, or flank pain, and change in clinical status with no other reason other than UTI. A negative urinalysis and no therapy supported the NNUTI diagnosis. Results. In 5 months, 29226 urine samples were evaluated. 401 patients were reported as NNUTI. Of these, only 5 (1.2%) patients met criteria for potential symptomatic UTI. Two patients treated for asymptomatic UTI were subsequently diagnosed with Clostridium difficile infection and renal failure, respectively. The second patient died of an adverse reaction to antibiotic therapy. No patient was adversely impacted by a NNUTI culture report. Conclusion. The new reporting criteria accurately reported the absence of a UTI in >98% of samples that had bacterial counts of <10 CFU/ml. Overtreatment of UTI has serious clinical consequences. Disclosures. All authors: No reported disclosures.

637Control of Methicillin-Resistant Staphylococcus aureus (MRSA) in Long Term Care is Possible While Maintaining Patient Socialization without Isolation

637. Control of Methicillin-Resistant Staphylococcus aureus (MRSA) in Long Term Care is Possible While Maintaining Patient Socialization without Isolation Becky Smith, MD; Susan Boehm, RN, BSN; Jennifer Beaumont, MS; Ari Robicsek, MD; Parul Patel, BS MT(ASCP), CCRP; Donna Schora, MT(ASCP); Deborah Burdsall, RN-BC, CIC; Kari Peterson, BA; Maureen Fausone, BA; Lance Peterson, MD; Infectious Diseases, Pritzker School of Medicine, University of Chicago, Chicago, IL; NorthShore University HealthSystem, Evanston, IL; Department of Medical Social Sciences, Northwestern University, Chicago, IL; Infection Control, Lutheran Home/Lutheran Life Communities, Arlington Heights, IL; Research, NorthShore University HealthSystem, Evanston, IL; Pritzker School of Medicine, University of Chicago, Chicago, IL

Urinalysis: The Microscopic Detection of Bacteria Does Not Predict Significant Bacteriuria

BACKGROUND/OBJECTIVES: Urinalysis (UA) is used to screen for urinary tract pathology including infection. The microscopic detection of bacteria, a component of themanual UA panel, predicts the presence of bacteriuria. We have found that reporting the presence of bacteria leads clinicians to treat for infection when other UA and culture findings do not suggest infection. We investigated the correlation between the microscopic detection of bacteria and a culture growing a potential pathogen. METHODS: 30,777 urine samples from May, July and October 2013 were analyzed. All samples were tested by manual UA, including microscopic detection of bacteria, and culture. No bacteria observed was considered a negative microscopic UA. Positive Results were reported as 1+, 2+, 3+ or 4+ bacteria present. A culture result of no growth or mixed flora was considered negative. Positive cultures were reported using isolate identification and quantitative counts to suggest potential pathogens. The sensitivity, specificity, and positive (PPV) and negative predictive values (NPV) were calculated. RESULTS: Therewere 2,865 positive and 27,912 negative specimens based on culture. Sensitivity, specificity, PPV and NPV for the microscopic detection of bacteria as a predictor of positive culture was 0.78, 0.51, 0.14 and 0.95, respectively. 13,661 (44%) false positive and 626 (2%) false negative microscopic UA reports occurred. 14% of patients with positive microscopic UA were confirmed by culture. 96% of patients with negative UA were confirmed by culture. If we considered a negative microscopic UA to include negative, 1+ or 2+ bacteria and positive UA to include 3+ and 4+ bacteria, the PPV increased to only 0.23. CONCLUSIONS: Overall, the microscopic detection of bacteria correlates with culture 55% of the time. The microscopic detection of bacteria within the UA panel does not accurately predict significant bacteriuria and should not be reported.