Incoming MPH Student Hadya Sow Chosen for Mayor's Office Internship
Incoming MPH student Hadya Sow has been chosen for the 2017 Mayor's Office Internship Program in the Office of Public Health and Safety. The program provides a unique opportunity for undergraduate and graduate students to learn and contribute to city government. Interns are assigned to different areas of the Mayor’s Office and work closely with all levels of local government, as well as the various departments that comprise the City of Indianapolis. Interns are assigned to substantive projects providing research, administrative, and staff assistance to city officials. They will also have the opportunity to learn about careers in local government, hear from city leaders, and experience the day-to-day operations. Join us in congratulating Hadya on this honor!
Joint Biostatistics Seminar
Join the biostatistics department on Friday, January 13 at 1:00pm in HITS 3139 for Dr. Giorgos Bakoyannis' presentation, "Semiparametric analysis of competing risks data for studies with double-sampling designs." His abstract: Competing risks data arise naturally in observational cohort studies and clinical trials. Frequently, outcome determination is based on imperfect, and usually inexpensive, diagnostic procedures which result in errors in outcome classification. This type of misclassification can lead to seriously biased estimates and thus invalid conclusions. In other studies with gold-standard but expensive diagnostic procedures, the total number of participants may be limited, due to cost. A useful approach to dealing with these problems is a double-sampling design where an imperfect and inexpensive outcome ascertainment procedure is being used for all the observed failures, and a gold-standard and expensive procedure is used to retrieve the true cause of failure for only a small sample from the non-censored cases. In this work we propose a maximum profile pseudolikelihood estimator (MPPLE) to analyze these data, under the semiparametric proportional cause-specific hazards model. We show that the MPPLE for the regression parameters is consistent and asymptotically normal. Additionally, the estimator for the covariate-specific cumulative incidence function is shown to be uniformly consistent and to converge weakly to a tight Gaussian process. Simulation studies show that the MPPLE works well with small to moderate sample sizes and small double-sampling probabilities. The method is not computationally intensive and thus it is useful in studies with large sample sizes. The method is illustrated using data from a cohort study of HIV-1 seropositive individuals in sub-Saharan Africa, where serious death under-reporting results in classifying many deceased patients as being disengagers from HIV care. In that study, a double-sampling design was used to deal with misclassification due to death under-reporting.
To ensure the protection of both the university and personal information that is maintained in IU systems, increased security has been implemented in various IU systems over the past several months. In June, the Employee Center began the use of the Two-Step Login methodology, and on September 1, the IUIE system is the next system to require it. IU currently utilizes the DUO system for the Two-Step Login. Please refer to the email sent on October 24 for more information on how to setup the Two-Step Login. Two-Step Login will be required in order to access your electronic W2 for 2016 after the first of the calendar year.
The DUO system allows for the use of a hardware token or a phone/mobile device as the second authentication method. Lync phones, which we have in most Public Health offices, are not valid for use as the second authentication method as they are activated by the same User ID/logon credentials used to access other IU systems. Alternatively, personal cell phones, mobile devices (tablets) and/or home phones (when working remotely) can be used to setup the required DUO notification process. In the past, FSPH provided hardware tokens to those with a valid business reason to access the few systems requiring this second level of authentication. This was done to eliminate the need for these individuals to utilize their personal phone for business purposes. But now, faced with the fact that more and more systems are requiring this type of logon process, we must address the issue of a widespread use of DUO. You can receive face-to-face support for Two-Step Login at an event on campus.
All faculty and staff at Public Health are encouraged to begin using a digital signature for email. Please follow the instructions in the link to obtain and install your digital signature: kb.iu.edu/d/bctk. These instructions must be followed carefully and may take some time to complete. Please note that the digital signature must be installed on any computer or device from which you wish to send digitally signed email. Some devices and systems have known compatibility issues: kb.iu.edu/d/aluf.
The purpose of using a digital signature is to help verify the identity of the sender, which helps prevent email spoofing and phishing attacks which are occurring with increasing frequency at IU.