Interview with Rohit P. Shenoi- winner of the 2024 Jess Kraus Award 🏆

Hi, Dr. Shenoi, congratulations to you and your team on winning the Jess Kraus Award in Injury Epidemiology! Would you please briefly introduce yourself and your co-authors to our readers?

I am Rohit P. Shenoi, MD, a Pediatric Emergency Physician at Texas Children’s Hospital and Professor of Pediatrics at Baylor College of Medicine. 

My research colleagues include:

Briana Moreland, MPH, a health scientist who was working at the Centers for Disease Control and Prevention in the study section on drowning prevention at the time of the study,

Jennifer L. Jones, MS, an epidemiologist at the Division of Emergency Medicine, Department of Pediatrics at Baylor College of Medicine at the time of the study,

Nicholas Peoples, MD, a 4^th year medical student at Baylor College of Medicine at the time of the study,

Elizabeth A. Camp, PhD, a research statistician at the Division of Emergency Medicine, Department of Pediatrics at Baylor College of Medicine, and

Ned Levine, PhD, an urban researcher who specializes in methodology with an extensive experience in geographical information systems and spatial analysis. 

How did you become interested in researching drowning? 

My interest in injury prevention began early in my career when I first studied injury patterns resulting from motor vehicle crashes in metropolitan Houston through a Texas Department of Transportation grant. The goal was to link data from motor vehicle crashes, EMS, hospital and fatality records and develop fact sheets which could be used for local injury prevention activities. This work helped set the stage for building a team of researchers and collaborating with a network of stakeholders.  

Water recreation is very popular in the Houston region given its warm weather, presence of many swimming pools, lakes, and rivers and its proximity to the Gulf of Mexico. Unfortunately, when the exposure to water increases, the chances of drowning also increase. I am a pediatric emergency physician and have treated many children who drowned. I observed that there are many knowledge gaps in the clinical management of drowned patients and in the epidemiology of drowning. The confluence of two factors: a high drowning burden in Houston and so many unanswered questions in drowning led me to my current research niche.

What motivated you to pursue this study?

There is no single source for obtaining the most current fatal and non-fatal drowning data. Most data sources that are used to study the epidemiology of drowning, such as CDC Wonder, utilize death certificate data. The drawbacks of these data sources are delayed data release, no information on non-fatal drowning and limited data (i.e., mostly demographics and body of water by ICD-10 codes).

Syndromic surveillance has been used to study the epidemiology of many infectious and non-infectious illnesses and injuries, such as those caused by firearms and opioid overdose. However, syndromic surveillance data has not been used to study the epidemiology of fatal and non-fatal drowning at the regional level. Since most drowning countermeasures are developed at the local or regional level, I believed that by assessing the accuracy of syndromic surveillance data in describing the epidemiology of drowning, we could provide local and state health departments the tools for performing real-time drowning surveillance utilizing an existing health information infrastructure.

Your study is based on data from the National Syndromic Surveillance Program (NSSP). Would you please tell us what the NSSP does, how it is operated, and what its major strengths and limitations are? 

The NSSP collects data from over 78% of emergency departments nationwide. Electronic health data are transmitted to a shared platform called BioSense, and public health agencies, such as the CDC and local and state health departments, can analyze data as early as 24 hours after a patient encounter. The system can detect, describe, monitor and assess the response to events of public health importance.

Captured visit information from ED and urgent care clinics includes free-text chief complaint, discharge diagnosis codes, and patient demographics. Diagnostic information is obtained using ICD-10 codes. There are syndromic surveillance definitions for many conditions or illnesses. The “CDC unintentional drowning v1” definition queries patient discharge diagnosis (ICD-10 codes) and chief complaint and discharge diagnosis information (text) to identify initial unintentional drowning encounters.

There are important strengths of the NSSP. First, it has an existing infrastructure and there are staff tasked with monitoring various illnesses and injuries at local and state health departments. Therefore, the data are readily available. Second, if the syndromic surveillance definition for a particular illness or condition exists, Biosense can be queried and surveillance data for that condition obtained very quickly. The rapid turnaround of data allows for disease and injury surveillance in real time. Researchers can access these data through data use agreements. Third, standardized data variables allow for comparison of injury burdens and trends between different regions or states or within the same region over time.

The system also has many drawbacks. First, the syndrome definition used might underestimate or overestimate SS visits because of variation in coding, reporting, and the availability of visit-level data between facilities or over time. Second, it is difficult to determine the intent of injury in SS data. ICD-10-CM codes are not included in all SS visits and even when ICD-10-CM codes are included in the discharge diagnosis field, not all codes are associated with a specific intent. Third, SS data lacks consistent information on patient disposition and it is difficult to determine if drowning visits are fatal or nonfatal. Fourth, SS data which are provided are deidentified and in aggregate. This does not allow SS data to be linked with other datasets. Fifth, duplicate cases may arise due to inter-facility transfer of patients. Finally, syndromic surveillance data cover about four-fifths of EDs in the country and are not nationally representative.

Your award-winning study was aimed at assessing the accuracy of the NSSP in capturing unintentional and undetermined intent drowning (UUID).  Would you please explain briefly the research method you used to achieve the study objective?     

We obtained deidentified syndromic surveillance data for the 8-county metropolitan Houston region (Texas Public Health Region 6/5 South) through a data use agreement with the Houston Health Department (HHD) and obtained ethical approval. We supplied the UUID drowning definition to the HHD who applied this to NSSP data. With the dataset obtained, we performed a 2-step process to identify all cases with UUID. First, we queried the dataset for UUID ICD-10-CM codes. Next, among records which lacked ICD-10 codes, we manually reviewed the chief complaint (CC) and discharge diagnosis (DD) texts of SS visits to identify UUID cases. We excluded cases that did not include submersion in water. True-positives were calculated by dividing the number of UUID cases identified by UUID ICD-10-CM codes and CC/DD review by the total visits captured by the SS definition. Thereafter, we performed descriptive statistics, explored injury trends and compared our results with NVSS data.

How is UUID captured by the NSSP? What are the reference criteria used in your study? 

The CDC developed and pre-tested the definition for “unintentional drowning v1” which is used for querying the NSSP. To this, we added ICD-10 codes for undetermined intent drowning and obtained all cases of unintentional and undetermined intent drowning (UUID). The unintentional drowning definition is a set of computer codes or syntax which we requested from the CDC and forwarded to the Houston Health Department when requesting SS data. Syndromic surveillance data describe emergency department and urgent care clinic visits and can be compared with hospital discharge data.

What are the key findings of your study? 

The accuracy of the syndromic surveillance definition for undetermined and unintentional intent drowning (UUID) is very high (98%). Syndromic surveillance can be used to monitor UUID burden and trends in real time at a regional level.

We also observed that the distribution of SS visits following drowning based on pediatric age group, male sex, race and Hispanic ethnicity are different from fatal drowning reports obtained from death certificate data.

What are the implications of your study results for drowning prevention? 

In regions with a high level of participation by medical institutions in the NSSP, syndromic surveillance data can be used to accurately and efficiently monitor fatal and non-fatal drowning burden (with some limitations) in real time among persons seeking medical care. It could also be used to monitor the effectiveness of drowning countermeasures in a region over time.

Since patient demographics among SS visits and death certificate data differ, this suggests that SS data and death certificate data provide information on two different populations. While death certificate data describe fatal drowning at the scene and along the continuum of care, SS data provide information on non-fatal and fatal drowning patients who sought medical care. Therefore, to conduct drowning surveillance at a regional level, it would be best to study SS and medical examiner data.

The emergency department is where clinical medicine and public health intersect.  As an emergency physician, could you share a few academic pearls with our graduate students and junior faculty aspiring to improve population health and safety? 

1. Clinical care allows us to apply a human face to injury. My role as a clinician in treating injured children has reinforced my efforts to prevent injuries and to advocate for the safety of children. While being very satisfying for me personally, it has kept me motivated in my research and advocacy all these years.
2. Find a mentor early in your career. I was fortunate to partner with my co-investigator, Dr. Ned Levine, in a completely different field - regional transportation safety, two decades ago. It helped me understand how governmental systems operate, urban planning, accessing and studying data systems, and networking with experts in different disciplines. These skills are transferable to other fields of study. A mentor can identify the protégé’s strengths and suggest opportunities for academic advancement.
3. Be cognizant of emerging illnesses and injuries and developing technologies and treatments. Think about how technologies and methods that were successful in the surveillance, management or prevention of one type of disease or illness could be applied to your field of study. In my case, I observed the use of syndromic surveillance in studying drug overdoses and applied it to drowning. 

Thank you for taking time to talk with me. I look forward to your seminar presentation and the award ceremony at Columbia University this fall.

Read Dr Shenoi's award-winning article: Using syndromic surveillance for unintentional and undetermined intent drowning surveillance in a large metropolitan area