Part-Time Biostatistician

Harvard Medical School

Boston, MA

Job posting number: #7076524

Posted: March 29, 2021

Application Deadline: Open Until Filled

Job Description

Job-Specific Responsibilities

This position involves working within a fast-paced interdisciplinary team on statistical analysis of survey data. Statistical analyses for both studies include, but are not limited to data reductions, causal modeling, machine learning, and measuring the effect of heterogeneity of treatment. Involves working with multiple, large datasets by formatting, merging, and sub-setting them to meet the analysis goals. A successful candidate must: make independent decisions about how to implement analyses most effectively; ensure that tables, figures, and text accurately reflect the results of the analysis; be able to understand statistical foundations behind models; prioritize tasks related to own responsibilities to meet goals and deadlines of the team; and work closely with Principle Investigator, Project Director, Senior Biostatistician, Data Manager and analysis team members at collaborating universities.

Studies Overview:
Analysis of EHR data on Kessler NIMH projects/Statistical analyses for two prospective observational studies using Veteran Health Administration (VHA) Electronic Health Records (EHR) to evaluate effects of key treatment decisions on suicide-related behaviors (SRBs; suicide deaths in the National Death Index or administratively-recorded suicide attempts (SAs)). The first study will estimate aggregate comparative treatment effects and then create Personalized Treatment Rules (PTRs) to determine the right treatments for the right patients. Statistical methods applied to large EHR databases to adjust for (“balance”) baseline differences in patients who received different types of treatment will be used to estimate aggregate treatment effects. It is feasible to clarify the effects of PCP decisions about how to treat common SRBs and about whether to hospitalize patients after SAs in order to answer the question “What works for whom?” by developing PTRs for these decisions using input from VHA EHRs. The second study carries out a pragmatic trial among psychiatric inpatients in the VHA healthcare system judged to be at high risk of post-discharge suicide (based on a validated prediction model). We will evaluate the effects of a scalable remote post-discharge intensive case management program (CLASP; Coping Long Term with Active Suicide Program) and also carry out a heterogeneity of treatment effects (HTE) analysis to determine how these effects vary as a function of patient characteristics and features of the post-discharge treatment environment. The primary outcome will be post-discharge suicide-related behaviors (including deaths due to suicide, opioid overdose, or other substance-related accidents; and nonfatal suicide attempts) in the 6 months post-discharge.


Basic Qualifications

Masters degree in statistics or related quantitative field and 5+ years relevant experience. Relevant graduate work may count toward years of experience. US Citizen Status required by US Army.

Additional Qualifications and Skills

Three or more years of R experience in a workplace preferred.

Additional Information

This is a grant-funded position for one year with the possibility of renewal.

This position requires a National Agency Check with Credit Check and Written Inquiries (NACIC) and fingerprinting conducted by the Security Division at USUHS to obtain Army STARRS level one security clearance. U.S. Citizen Status required by US Army. This position also requires a US Office of Personnel Management (OPM) check and a criminal check conducted by the VA (Veteran’s Administration).


Harvard Medical School strives to cultivate an environment that promotes inclusiveness and collaboration among students, faculty and staff and to create new avenues for discussion that will advance our shared mission to improve the health of people throughout the world.


Apply Now

Please mention to the employer that you saw this ad on BiomedJobs.com