You are currently viewing Case Study: Heat Map for Super Refractory Status Epilepticus

Case Study: Heat Map for Super Refractory Status Epilepticus

I have grown many businesses by using consistent digital marketing tactics. However, the most creative and unique thing I have done using digital marketing was come up with a unique solution to an uncommon problem.

My client was a pharmaceutical company that developed a drug to treat the rare condition Super Refractory Status Epilepticus. Super-refractory status epilepticus is an ongoing seizure over a 24 hour period that is non-responsive to usual seizure meds.

The problem was two-fold. First, the condition is rare, so finding incidents was also a rare occurrence. Second, the window for treatment was small. The medication had to administer within that 24-hour time frame to be effective. The condition is so rare that a good number of healthcare providers are not aware of the state until they are face to face with an incident. We knew that when healthcare providers were faced with a case, they turned online resources for help.

My solution for identifying incidents was to monitor search engine traffic. First, we ran paid search ads over 30 days to determine the mean impressions for all relevant DMA’s. Then we monitored impressions via each DMA daily, checking for impression numbers that lie outside of the normal standard deviation of the mean for each DMA.

Background:

The client was a pharmaceutical company that has developed a drug to treat the rare condition Super Refractory Status Epliepticus. Super-refractory status epilepticus is defined as status epilepticus that continues or recurs for more than 24 hours after the onset of anesthetic therapy, including those cases where status epilepticus recurs on the reduction or withdrawal of anesthesia. It is an uncommon but essential clinical problem with high mortality and morbidity rates. (medscape.com)

The Problem:

  • We would like to know what is the best statically sound way to find a 30 day mean for each DMA, especially we take into account.
  • We know that each group of keywords carries more weight than the other. For example, If there is an uptick in previous day impressions for Seizure keywords in Denver, CO, but there is no uptick in the previous day’s impressions for SRSE keywords OR Disorder keywords, there probably isn’t a real incident. We had to create a weighted average in which we give SRSE more weight. How do we do this?
  • Once we have a mean number of daily impressions for each DMA, per each set of keywords, we would like to know HOW to identify a significant increase in impressions. In other words, how do we define significance?

Project Goals:

SRSE is so rare that a good number of doctors, nurses, and pharmacists are not aware of the condition until they are face to face with an incident. Once the condition is identified, usually, the only way to treat an SRSE patient is a medically induced coma to avoid cerebral damage. Our job is to identify possible incidences within the US as quickly as possible via paid search so that the condition can be treated within its early stages.

Methodology:

We found that the best tool at our disposal to identify an SRSE incident is to monitor search engine traffic. Our logic is that in the scenario in which a healthcare provider encounters an SRSE incident, they are more than likely to seek diagnosis and treatment information online with a search.

During our initial analysis, we have found there is a negligible amount of search traffic for SRSE terms daily. With this in mind, we conducted keyword research that allowed us to find every possible search query that, in theory, related directly to an SRSE search. Our keyword research took into account the full spectrum of search terms. We created an index categorizing these terms into groups based on relevance and assigned each group a weight according to that relevance. These groups included: Brand, Disorder, Seizure/Treatment, and SRSE. We reviewed our findings with the pharmaceutical team to vet out our research and ensure our project match their specifications and goals.

  • Disorder – keywords related to status epileptics. (status epilepticus is the condition before it reaches the critical “super” phase.)
  • Seizure – keywords that are most likely being searched by family members trying to figure out what is wrong with their loved one.
  • SRSE – super refractory status epilepticus keywords. These keywords are the most valuable since they are the MOST relevant.
  • Brand – Brand keywords are keywords that identify the pharmaceutical company, Sage Therapeutics.

Super-refractory status epilepticus is defined as status epilepticus that continues or recurs for more than 24 hours after the onset of anesthetic therapy, including those cases where status epilepticus recurs on the reduction or withdrawal of anesthesia.

The second phase of this project is to identify what is “normal” SRSE traffic for each significant US DMA. To do this, we created paid search campaigns using the keyword research we conducted and uploaded these campaigns to Google Ads. Our ads would be triggered anytime someone searches for SRSE related keywords anywhere with the continental US using paid search. The key metrics we kept an idea on were our impressions and average position. We want to make sure that our ads would be triggered at a consistent rate without falling to the second page.

We identified regular traffic by running paid search campaigns for 30 days and then calculating a mean amount of daily impressions for each US DMA. The mean gave us a good idea of what “normal” search volume is for each region. If a region’s search volume fell two standard deviations above the mean for a particular DMA, then that would indicate a possible SRSE incident.

The third stage of the project was to finally launch campaigns and pull daily morning reports for the previous day. We set up automated macros to highlight higher than average impressions. These reports were pulled and sent to the pharmaceutical team daily.

Summary:

  1. Our first goal is to identify mean daily impressions via each geographic DMAs over 30 days for SRSE keywords.
  2. Once we have a mean daily impression number for each DMA for each set of keywords, we began pulling reports from the previous day that show us if there is a significant increase in impressions within each DMA.
  3. If there is a significant increase, we flagged the DMA and sent the report to the SAGE team. The SAGE commenced scouting the DMA for possible SRSE incidences.

Leave a Reply