From the Director
![DUKE.RESEARCH.NIGHT.03 (1)](http://news.medicine.duke.edu/wp-content/uploads/2013/06/DUKE.RESEARCH.NIGHT_.03-1.jpg)
![BB - ALL](http://news.medicine.duke.edu/wp-content/uploads/2014/05/BB-ALL.jpg)
![BB - Winners](http://news.medicine.duke.edu/wp-content/uploads/2014/05/BB-Winners.jpg)
![ST Set Up](http://news.medicine.duke.edu/wp-content/uploads/2014/05/ST-Set-Up.jpg)
![ST Team](http://news.medicine.duke.edu/wp-content/uploads/2014/05/ST-Team.jpg)
What Did I Read This Week?
Diagnosis of Childhood Tuberculosis and Host RNA Expression in Africa
http://www.nejm.org/doi/pdf/10.1056/NEJMoa1303657
Submitted by Aimee Zaas, MD
[/box]
Why Did I Read This: Went to check out the NEJM table of contents to find a good article for WIRTW. The review article on LYME DISEASE in this week's issue is worth reading as well (it's finally spring! Everyone thinks they had a tick bite! Do I give them doxy?). However, I really wanted to see this article where they use host expression for diagnosis of infectious diseases, as this has been a huge area of focus here at Duke for the past several years. What is the Reason to Evaluate Host Expression for Diagnosis of Infectious Diseases? : In the situation presented by the authors, (childhood Tb in Africa), it is a COMMON (500,000-1 million cases diagnosed annually) disease that is difficult to diagnose as many children with Tb are smear/culture negative, and since the clinical symptoms overlap with many other common diseases of childhood in the developing world, both over and under diagnosis are common. IGRA's don't distinguish latent versus active disease, and can be unreliable in compromised hosts. Therefore, a new diagnostic method is imperative. This theme is the same for many infections – either finding the pathogen is hard, our clinical definitions are hazy or we can't tell the difference between colonization and infection. Evaluating host response is one way of getting past the limits of pathogen based diagnostics. What Did the Authors Do? After getting approval from all the relevant IRBs, they recruited African children from 3 countries with a high burden of Tb. They formed a "discovery" cohort of children from Malawi and S Africa and a validation cohort in Kenya. Discovery and validation cohorts are a standard procedure in genomic studies like this one, so that you can show that your data can be replicated in a new set of patients. To be considered for the study, you had to be under age 15 and meet certain clinical criteria: cough/fever/wt loss lasting > 2 weeks, pneumonia not responsive to antibacterials or "suggestive clinical findings" or close contact with an adult who had active Tb. CXR, sputum and other clinically relevant body fluid specimens were obtained for Mtb smear/culture; IGRA was performed as was CRP and HIV testing. Follow up at 3 months determined if those without Tb remained disease free and also the response to therapy in those with Tb. Cases were defined as "definite Tb" for culture positive subjects, and for culture negative subjects, they were defined as "highly probable", "probable" or "possible" Tb. A positive IGRA indicated "latent" disease and "no Tb" was the label for subjects who remained Tb free at 3 months without antiTb therapy.1356 children were screened for the discovery cohort, with ultimately 364 included. 114 had culture confirmed Tb, 157 had diseases other than Tb and 57 had latent Tb. The authors did not evaluate children with culture negative but clinically likely Tb (the "highly probable" "probable" "possible" group) for the discovery cohort. For the validation cohort, 1599 children were screened and 157 selected for the study. In this group, 64 had no Tb, 35 had culture + Tb, 9 had latent Tb and 44 had culture negative Tb (8 "highly probable" 19 "probable" 17 "possible"). How did they test RNA: RNA Paxgene tubes were drawn on each subject. These are special blood draw tubes that have a means of "freezing" cellular expression so that when you extract RNA, you know exactly what the cell was doing at the time you drew the sample. The samples later had the RNA extracted and hybridized to Illumina microarrays. They then looked for genes having differential expression (eitherlog2 of 0.5 up or down) between Tb and no Tb – and whittled this list down to the minimum number of genes possible using an algorithm called an elastic net. This gene list became used for their "risk of Tb" score, calculated by evaluating the difference in the total intensity of up regulated genes minus intensity of down regulated genes. What did they find? In the training set (comprising 80% of samples from the discovery cohort), they identified 409 transcripts that were differentially expressed between tuberculosis and other diseases and 3434 transcripts that were differentially expressed between tuberculosis and latent infection. Using variable selection to identify the smallest number of transcripts that distinguished each group, they found that 51 transcripts distinguished tuberculosis from other diseases and 42 distinguished tuberculosis from latent infection. These transcript sets were used to generate a risk score for each patient in the test set (comprising the remaining 20% of samples from the discovery cohort) that distinguished tuberculosis from other diseases (sensitivity of 78% and specificity of 74%) and that distinguished tuberculosis from latent infection (sensitivity of 96% and specificity of 91%) In the validation cohort, The risk score discriminated culture-confirmed tuberculosis from other diseases in patients with or without HIV infection with a sensitivity of 82.9%and a specificity of 83.6%. Among patientswith negative cultures who were treated for tuberculosis, the risk score identified 62.5% of those in whom tuberculosis was highly probable, 42.1% of those in whom it was probable, and 35.3% of those in whom it was possible. The risk score did better than the pathogen based GeneXpert MTb test performed on respiratory samples (this is a gene probe test to identify presence of MTb). They then performed a sensitivity analysis, estimating how the gene score would perform in groups with 10%, 30% and 50% pretest probability of having MTb. As expected, the test performs better for finding Tb when there is a higher pretest probability. Overall, however, this test had better NEGATIVE predictive value than POSITIVE predictive value, so perhaps has a more defined role in "ruling out" disease. Limitations include the cost, and a much cheaper way of processing RNA needs to be developed before this could be used in resource limited settings. When looking at the gene list, they do not make any attempt or any claims to say that the genes they found have biologic plausibility to be related to Tb, although 3 defensins and one IFN related gene are represented. What can we conclude: This is a strong "proof of concept" study using solid methodology to look for novel ways to diagnose Tb in a setting where our traditional diagnostics are limited. It is certainly not ready for "prime time" both because it needs to be validated again in a prospective manner and also because cost is a huge factor and this test would not be practical until we can easily and cheaply measure RNA expression. How does this apply to your practice? Well, immediately it does not apply to your practice, even if you are headed to a Tb endemic area next year for a global health elective. However, the concept of host based diagnostics for infectious diseases is growing (think of procalcitonin in the simplest scenario). As expression arrays and other 'omics technologies become cheaper and more automated, we will see these kinds of tests in clinical practice and will need to be able to interpret the results.![WDITTW Graph](http://news.medicine.duke.edu/wp-content/uploads/2014/05/WDITTW-Graph.jpg)
The "Clinic Corner"
(submitted by Alex Cho, MD )
This week’s Clinic Corner is pretty simple – just an announcement of something to look forward to: The spring Ambulatory Town Hall meetings are coming up in a few weeks, scheduled for Tuesday May 27 over the noon hour. Come and catch up on the latest in your continuity clinic, and share your thoughts and suggestions. DOC and PRIME will be in DN 2002/3, and Pickett Rd in the MedRes Library. Don’t miss it! And thanks to everyone who completed the inaugural Learner Perception Survey (LPS) Snapshot clinic survey that Duke IM grad and first-year GIM fellow Denise Duan-Porter developed with help from the VA’s national LPS group. (Congrats to the Kerby Stead Society for raising $250 for Senior PharmAssist/DOC Patient Fund by having the highest completion rate!) The results have been shared with the clinic leaders and will help inform ongoing work at the clinics to improve patient care as well as the resident experience [divider]From the Chief Residents
Grand Rounds
Dr. Souha Kanj – Greenfield Visiting – Infectious Diseases Topic: Infections in the Middle EastNoon Conference
Date | Topic | Lecturer | Time | Vendor | Room |
5/5 | MKSAP Mondays | Chiefs | 12:00 | China King | Med Res Library / 8262 |
5/6 | Clinical Reasoning I | Hargett | 12:00 | Pita Pit | 2002 |
5/7 | IM-ED Combined Conference - COPD | Kussin | 12:00 | Subway | 2002 |
5/8 | Clinical Reasoning II | Hargett | 12:00 | Domino's | 2001 |
5/9 | Chair's Conference | Chiefs | 12:00 | Chick-fil-A | Med Res Library |
From the Residency Office
SAFETY ATTITUDES QUESITONNAIRE ABOUT TO START!
What is it: Culture of Safety Survey, second full cycle for DUHS When is it: Survey runs May 5-May 30. Who does it: All ACGME program members will be included (if at least 8 members); other clinical departments throughout the health system also doing survey How is it done: Participants will get an email from support@pascalmetrics.com, with subject “DUHS Safety Culture Survey from Pascal Metrics”; in the body of the email, the target/referent for the survey is listed. For GME it is the GME Program (e.g. GME-Medicine-Internal Medicine). Last cycle, GME participation rate was 71%. We are looking for at least 80% response rate from each program! Your input is very important.Residency Program Evaluation by Residents
All Internal Medicine residents have been assigned, via MedHub, an evaluation of the residency program. Please complete the evaluation no later than May 23rd. Your feedback is extremely important to us and the information that we are able to gather via this evaluation will help tremendously as we continue to investigate ways to improve in all areas. In response to the results we received from our 2012-13 ACGME Residents Survey, we would like for you to pay particular attention to the questions concerning duty hour violations, as we continue to examine how to improve these issues within specific services. Thank you in advance for your time and attention to this! Please feel free to contact Jen Averitt if you have any questions.Duke Hospital Medicine Social
If you are considering a career within Hospital Medicine, the following is an opportunity to meet with our hospitalist team to discuss opportunities in greater detail.- What: Hospital Medicine Interest Meeting
- Date: May 19, 2014
- Location: Nosh Café Trent Semans Center
- Time: 5:00pm
Information/Opportunities
Hospitalist Practice Opportunity 5-2014Upcoming Dates and Events
- May 24: Housestaff Party - Elodie Farms
- May 30: Program pictures, Trent Semans West Steps, 9:15
- May 31: SAR Dinner, Hope Valley CC
- June 3: Annual Resident Research Conference
- June 6: Serve dinner at the Ronald McDonald House
Useful links
- https://intranet.dm.duke.edu/influenza/SitePages/Home.aspx
- http://duke.exitcareoncall.com/.
- Main Internal Medicine Residency website
- Main Curriculum website
- Ambulatory curriculum wiki
- Department of Medicine
- Confidential Comment Line Note: ALL submissions are strictly confidential unless you chose to complete the optional section requesting a response