Ghost in the Machine: The Emotional Gravity of Conducting Mortality Research

An epidemiology graduate student recently demonstrated to me that analyzing vital statistics carries occupational hazards. Reviewing overdose mortality data, the student was viscerally reminded of childhood experiences precipitated by their father’s substance use and undiagnosed mental health issues. In between neat rows of structured data, the father’s death reasserted itself, along with the family’s trauma. Although the student’s passion for public health stemmed from these very experiences, the scholastic environment had not allowed for expression of this identity. Therefore, when distressing feelings arose, there was no antecedent. It was a missed opportunity to support a trainee.

The emotional gravity of working with mortality affects seasoned researchers too. With a few taps, I can securely access tens of millions of death records from my phone. These data weigh heavy in my pocket as I go about my day. After two decades of working with these data, when I duck into a car, I wonder how soon I may end up as a death record in a row. Like many others, I can vividly remember the first time I saw a friend’s death record in a data set. (Although ostensibly anonymized, the circumstances were clear.) I caressed the row with my mouse; I left the window open on my desktop for days, feeling his presence emanating through pixels. I do not think these experiences make my research any less objective. Rather, they hold me accountable to a higher power.

I take regular walks through two historic cemeteries flanking my neighborhood. I do not know a soul there; most of the people died before people of my ethnic background were even allowed into the country. No causes of death are mentioned on the tombstones, yet the decorative filigree, monumental regalia, family plot spacing, and epitaphs offer communion across centuries. Why do I feel a sense of calm when walking out of the cemetery but feel unsettled when shutting down my computer after days of analysis? I have come to realize that highly structured death data are vulnerable to emotional truncation. The encoded and medicalized encapsulation suits our surveillance needs, their parsimony implying that these are the only factors that matter. They are not.

Protection of research participants is a standard concern, but less attention is paid to the emotional well-being of analysts. Compassion fatigue and secondary researcher trauma have been described extensively in qualitative research. Sikic Micanovic et al. provide a concise review, ¹ and Kumar and Cavallaro present a useful framework. ² Qualitative research emphasizes the investigator’s viewpoint and orientation toward the subject matter because the research paradigm fully accepts that such acknowledgment can enrich interpretation. However, the dispassionate façade of quantitative research blinds us to analysts’ feelings: paradoxically, it is assumed that thousands of death records exert no mental toll. As my personal vignettes suggest, vital statistics data also carry emotional weight.

Emotional danger is defined as negative “feeling states” induced by the research process. This means not just feeling uncomfortable, but also manifesting distress that affects interpersonal relationships. ¹ By ignoring emotional impacts, we may be compromising our staff and results. Beyond the beguiling cleanliness of structured data, every data point embodies heartache. However, rows upon rows can lead to inurement. How do we retain our humanity? Qualitative researchers have suggested journaling the research experience and engaging in structured debriefing sessions. ³ Incorporating researcher well-being should be part of research design; we must dispense with the practice of masking emotions for the sake of projecting a professional posture. Personally, I keep photographs on my desk of loved ones who have passed. Sometimes, I light a candle and ask them if I am representing them with fidelity. This is one of my rituals, and I welcome you to share yours.