Page 77 - ISMCON souvenir 2021

ISMSCON - 2021

Results: A total of 1014 responses were followed up and only 407 of them consented to participate in
the study.The Mean age for >6mnths - ≤24 months, was found to be 15.76 ±4.81 months . Overall the
mean age was 20.91 ±12.86. months.. Most of the children (female & male) eats 3 times a day. Only 14%
children were eating four or more than four times a day. 45% children (171) were taking more than 4 food
groups & 52 % children were taking minimum 3 or less food groups. Conclusion: Our study highlights
the poor practice regarding hygiene, food diversity and frequency of food to the SAM children residing in
the state of Jharkhand.

OS46: MODELING OF SEMI-COMPETING RISK DATA IN
PRESENCE OF CENSORING

Rajashree Dey, Soutik Halder, Jitendra R. Gawde , Sunil K. Yadav, Sanjay D. Talole &
Atanu Bhattacharjee
Section of Biostatistics, Centre for Cancer Epidemiology, Tata Memorial Centre, India
(Email id-rajashreeedey@gmail.com)
Abstract:
In biomedical research involving time-to-event data, individuals may be liable to multiple possible
outcomes. When an individual experience more than one event in the follow-up process, this gives rise
to multiple failure time data. Here we consider a semi-competing risks framework in the presence of
interval-censoring and informative loss-to-follow up, where an individual may experience two distinct
types of events – terminal (e.g., death) or non-terminal (eg. Cancer relapse). where the terminal event
censors the non-terminal event but not vice versa. In the modeling of such data, Accelerated Failure Time
(AFT) models, an alternative to the traditional multiplicative Cox model that places emphasis away from
the hazard function can be used for the analysis of time to event data to estimate the effects of covariates
on acceleration/deceleration of the survival time. The statistical inference is based on a nonparametric
Bayesian approach that uses a Dirichlet process prior to the mixing distribution. An efficient computational
scheme, based on the Metropolis-Hastings-Green algorithm, was developed and implemented in the
SemiCompRisks R package. In this R package, the model was developed using continuous covariates
in scrdata. We have utilized categorical covariates on simulated data of n=5000 derived using the data
framework of ACT study to test the illness–death model. The deviance information criteria (DIC) for AFT-
LN vs. ADT-DPM model was found to be 37530 vs. 23409, and the corresponding Log–pseudo marginal
likelihood (LPML) was -16164 vs. -11475. This suggests the superiority of AFT-DPM over AFT-LN model.
Keywords: Bayesian survival analysis, illness-death models, interval-censoring, semi-competing risks

OS47: Trends and future burden of cervix cancer incidence in
Delhi using age-period-cohort regression models

Rajeev Kumar Malhotra , Nalliah Manoharan , SVS Deo 3
1
2
1 2 Scientist, Delhi Cancer Registry Dr BRA IRCH AIIMS, 3Professor and Head, Department of Surgery

Abstract
Introduction: Cervix cancer[CC] is ranked fourth most common women cancer globally for incidence as
well as for mortality. It is a preventable cancer and contributed nearly 7.0% of total new women cancer
worldwide.

Objective: To assesses the temporal trend of cervix cancer in Delhi using past data (1990-2014) and
projection of new cervix cancer cases using regression methods in year 2030

CONFERENCE SOUVENIR 75