Statistical Methods and Analysis of Environmental Health Data

The following was written by SukhDev Mishra,Ph.D., Division of Bio-Statistics, National Institute of Occupational Health, Indian Council of Medical Research, Ahmedabad(India)

group shot

Statistical Methods and Analysis of Environmental Health Data Workshop group.

I was fortunate to attend the SAMSI workshop on Statistical Methods and Analysis of Environmental Health Data last week in Mumbai. It focused on various topics related to the statistical analysis of environmental health data, some of which discussed latest methodological development in this field, particularly during the first day’s opening lecture from Professor Joel Schwartz.

Time series data has proven to be critical in the assessment of systematic impact of environmental factors on human health. Professor Francesca Dominici, a researcher with significant contributions in this area was a very dynamic and enthusiastic co-leader for this workshop. She discussed in length the statistical principles and assumptions of multi-site time series analysis along with careful interpretation of such data. Due to technological advances and regular measurement availability, time series data could be accessed and easily analyzed with the techniques elaborated by Professor Dominici, which will be integral to the success of my future studies.

Working Group 5 - Gene x Environment Interactions

Working Group 5 – Gene x Environment Interactions

The Gene x Environment Analysis & Epigenetics lecture taken by Professor Bhramar Mukherjee provided very useful information on interaction/additive and multiplicative models citing practical applications in area of environmental health that she developed. Her very creative way of teaching, blended with great sense of humor, kept us engaged so much so that we wouldn’t blink for a second.

Spatial statistics is a critical part for environmental health data, so it was helpful to have the basics covered by Dr. Safraj Shahul Hameed and Dr. Brian Reich well. Professor Donna Spiegelman presented a wonderful talk on measurement error starting from statistical notations to complete logit function (being a statistician ….I always love this part J ). She put great effort explaining Regression calibration method for MS/EVS and algorithms. Interesting talk!

Working groups were engaged in different exercises that included working on different problems/real data sets generated through various participants and coming up with new analysis and interpretation of data. I worked on Exposure Modelling of Ambient and Household Air Pollution for Acute and Chronic Health Effects. I enjoyed working with my fellow WG colleagues- Kalpana Balakrishnan, Santu Ghosh, Donna Spiegelman, Kevin Lane, Joel Schwartz, Sourangsu Chowdhury , and Poonam Rathi. Fine scientific arguments during the process of analysis were the crux of our exercise; thanks to Joel, Kalpana, Donna and Kevin especially.

This is no way a comprehensive description of this workshop, just my thoughts. I would also like to record here that I learned from each and every speaker and fellow participant. It was a gathering of great scientific minds and very inquisitive researchers. My understanding is that one of SAMSI’s objectives is to foster a culture of collaborative research among Indo-US researcher in area of public health; and I could see that coming true as we collectively discussed ideas on how to continue our work in mutual scientific engagement. I hope these efforts result in great scientific endeavors in coming time for environmental health priorities.

People drinking tea during a break

Enjoying afternoon tea.

One of the unique features of this workshop was meticulous planning by the team of organizers, be it scientific contents or overall execution by Professor Richard Smith, Professor Sujit Ghosh, Professor Francesca Dominici, and Ms. Krista Coleman whose scientific management and interaction with participants was very encouraging.

My working experience mainly includes working in pharmaceutical industry earlier, as biostatistician, and I consider myself a beginner in environmental health. This workshop has helped me to gain more scientific perspectives in this area by leaps and bounds.

This kind of knowledge sharing exercises may prove very helpful for researchers in the area of statistics and epidemiology to address India’s most pressing public health needs. Thank you SAMSI, Harvard, ISI-Kolkata and all of the other participating organizations for such a wonderful experience!



Topics in Probability: a VI-MSS Workshop

Science Across Virtual Institutes (SAVI) is a NSF program that facilitates collaborations among researchers from different countries.  One of the awards under this program is the Virtual Institute for Mathematical and Statistical Sciences (VI-MSS).  This award connects two U.S. NSF funded research institutes (SAMSI and ICERM) with several leading research institutes in India (Indian Statistical Institute (ISI), Chennai Mathematical Institute (CMI), the Indian Institute of Science (IIS), the Institute of Mathematical Sciences (IMSc) and the Tata Institute of Fundamental Research (TIFR)).  VI-MSS provides funding to U.S. graduate students, post doctoral fellows and faculty for research visits to partner institutes in India. Another key objective of VI-MSS is to sponsor joint workshops that have the potential to lead to collaborative research between scientists in the two countries.  The workshops are co-sponsored by the Indian federal research funding agency, Department of Science and Technology (DST).

group photo at the CMI workshop

Participants from the SAMSI-SAVI Workshop on Topics in Probability at Chennai Mathematical Institute, Siruseri, Tamilnadu, India

Topics in Probability was one such workshop that was held at CMI on December 18-20, 2012.  The workshop was co-organized by me and the director of CMI, Professor Rajeeva Karandikar.  CMI, located in Siruseri at the outskirts of Chennai was founded in 1989 as part of the SPIC science foundation and is a premier research institute in Mathematical Sciences in India.  The workshop brought together leading researchers in Probability Theory;  from U.S. Universities (Professors Richard Bass, Sandra Cerrai, Tom Kurtz, Ramon Van Handel and Mathukumalli Vidyasagar);  and  from several top Indian research institutes (Professors Siva Athreya, Vivek Borkar, Arup Bose, Manjunath Krishnapur, Krishanu Maulik and Anish Sarkar).  A broad range of topics were represented, including Random Matrices, Diffusion Processes, Percolation Theory, Infinite Dimensional Stochastic Analysis, Stochastic Partial Differential Equations, Random Graphs, Limit Theorems, Information Theory, Discrete Probability Models and Elliptic PDEs with applications in Probability.  Approximately fifteen graduate students, both from U.S. and India, participated in the workshop.  Talks were spread over three days with 3-4 talks each day giving ample opportunities for interaction and discussions.  As a result of a small number of talks per day, the pace of lectures was leisurely and the environment was relaxed and informal with frequent questions and extended discussions.