1 - (SYM 91) How Often Is Often Enough? Determining the Optimal Sampling Frequency When Assessing Suicidal Thinking

In the past decade, it has become increasingly clear that suicidal ideation (SI) varies considerably over short periods of time (e.g., hours, days) and that SI should therefore be assessed using methods like Ecological Momentary Assessment (EMA) that can capture this variability as it happens. EMA is particularly useful because short surveys can be sent (e.g., via smartphone) throughout the day at multiple times to get a picture of what real-time variability in SI looks like. One of several open questions around the use of EMA to study SI is the ideal number of times to assess it throughout the day. It could stand to reason that assessing SI as many times as feasible is the optimal solution. However, in practical terms, participants are willing and able to only answer a certain number of surveys each day. Thus, the goal of this study is to examine the increased information gained with each additional survey completed each day when assessing suicidal thoughts.

Two EMA studies were conducted with adolescents in the 4 weeks following discharge from inpatient hospitalization for acute suicide risk. In both studies, we assessed in-the-moment SI up to six times each day using a three-item measure commonly used in EMA research. The measure consisted of the desire to die by suicide, the intent to die by suicide, and the ability to keep oneself safe. We compared the information obtained on a daily level from all assessments (i.e., up to 6) against the information obtained by only three assessments or only one assessment (chosen at random) each day.  

We found that the daily mean and standard deviation for SI obtained when examining 3 or 6 daily assessments significantly differed from analyzing a single daily survey, suggesting that assessing SI only once each day misses valuable information about the daily occurrence of suicidal thoughts. However, the daily mean and standard deviation for SI did not differ when comparing 3 to 6 assessments. The same pattern was observed when examining person-level variables. Similarly, the ICC and RMSSD for suicidal thoughts differed significantly when comparing 1 assessment per day to 3 or 6 assessments, but 3 vs. 6 assessments did not yield significantly different information.

These findings indicate that more than one assessment per day is necessary to capture the daily phenomenology of suicidal thoughts among adolescents. However, in balancing the burden of frequent assessments with the value of the information we receive, six assessments per day may not be needed.