Surat Municipal Institute of Medical Education and Research
Contact: Jayeshkumar Kanani
Country: India
Award Category: Community Uplift
Tell how you've used SAS to have a positive impact on your community.
I have leveraged my SAS skills to positively impact my community. I acquired these skills through self-directed learning on SAS.com, successfully completing the SAS Base, Advanced, and Clinical Trials Programming exams with a perfect score of 1000/1000 on the Base exam.
In my role as an autopsy pathologist, I investigate the causes of death, particularly in non-natural cases. My primary focus within non-natural deaths lies in understanding suicidal behavior with the aim of contributing to prevention efforts. While investigating these cases, I encountered significant challenges in identifying the mental and environmental factors that may have contributed to suicidal attempts. This is due to the limited availability of information regarding an individual's state prior to their death. However, I recognized the potential for utilizing temporal patterns within suicidal deaths to identify periods of heightened vulnerability for suicidal ideation and attempts. A key challenge in this work is accurately determining the timing of suicidal attempts, as information about the individual's state prior to death is often unavailable.
Most existing studies rely on retrospective data from government sources, often using the declared time of death for analysis. However, this can be inaccurate as the time of death may be delayed (e.g., upon hospital arrival) or declared by a physician after examination. These inaccuracies can hinder the identification of critical periods for suicidal ideation and attempts, crucial for effective prevention strategies.
To address this limitation, I developed innovative approaches to pinpoint the precise time of suicidal attempts. In poisoning cases, this information can be obtained from medical records, and in railway accidents, from train driver reports. For hanging cases, I employed pragmatic methods to gather information on the time between when the deceased was last seen alive and when they were discovered. This involved:
• Interviewing relatives to ascertain the last time the deceased was seen alive.
• Thoroughly investigating witness statements, reviewing CCTV footage, examining call details, and analyzing social media activity.
• Examining stomach contents to narrow down the time window.
By carefully considering these factors, I was able to estimate the time of suicide attempt with greater accuracy. critical periods. Identifying the peak times of suicide attempts can inform preventive strategies. By understanding when suicide attempts are most likely, resources, including healthcare professionals and crisis hotlines, can be allocated more efficiently during high-risk times to reduce the likelihood of transitioning to completed suicides. Highlighting specific time patterns can help to reduce the stigma associated with suicide. This information underscores that suicidal thoughts and behaviours can be time-dependent and not solely reflective of an individual’s overall mental health. Data on the timing of suicide attempts allows policymakers, healthcare providers, and researchers to make informed decisions and develop evidence-based strategies to prevent suicide.
My research highlights the critical importance of accurate temporal data in understanding and preventing suicidal behavior. By utilizing SAS and employing these innovative methods, I have contributed valuable insights to enhance community health and safety initiatives.
What SAS products are you using and how are you using them?
For this research, I primarily utilized SAS On Demand for Academics, a valuable resource that provided access to the necessary SAS software.
• Data Import and Preparation: I employed the PROC IMPORT statement to efficiently import forensic datasets. Subsequently, I leveraged the PROC FORMAT procedure to standardize time and date variables, ensuring consistent and accurate data representation. To further prepare the data, I extensively utilized data step procedures, incorporating a combination of character and numeric functions to standardize variables such as age and sex into meaningful groups.
• Data Exploration and Visualization: I effectively explored the data using a variety of SAS procedures. The PROC SGPLOT and PROC SGPANEL procedures proved invaluable for creating insightful graphical representations, enabling visual exploration of trends and patterns within the data.
• Statistical Analysis: I conducted rigorous statistical analyses using a combination of PROC SQL and PROC FREQ. The PROC FREQ procedure, coupled with the chi-square test, allowed me to investigate the association between different variables and identify statistically significant relationships within the data.
What was your most surprising discovery about your work?
One of the most surprising discoveries of my research was the significant diurnal variation in the timing of suicide attempts.
• Day/Night Distribution: A substantial majority of suicides, approximately 86.5%, occurred during daylight hours. This initial finding highlighted the importance of considering the time of day as a crucial factor in understanding suicidal behavior.
• Morning/Afternoon/Evening/Night Distribution: Further analysis revealed a more nuanced picture. Suicide attempts were most prevalent in the afternoon (30.7%), followed by the evening (27.4%) and morning (23.7%). Notably, these patterns exhibited variations across different subgroups. Males were more likely to attempt suicide in the afternoon, while females showed a higher incidence in the evening.
• Hourly Distribution: Examining the data on an hourly basis provided even greater granularity. Suicide attempts peaked between 4 PM and 5 PM (7.91%), followed by the period between 5 PM and 7 PM (6.98%). These hourly peaks varied across gender, age groups, marital status, and even by the method of suicide. For instance, young adults tended to attempt suicide around 3 PM, while middle-aged adults showed a peak around 6 PM.
