Building and Benefiting from Living Memory Home

Principal Investigators: 
Holly G Prigerson
Sponsored by: 
NIH/NIMH
Research Description: 

Living Memory Home

The study is divided into two phases: pilot and formal study. The overall aim of the study is to

examine the relationship between the use of the LMH – a web-based application to honor the deceased

person’s memory – and symptoms of Prolonged Grief Disorder (PGD). The LMH is based on the idea

that maintaining a continuing bond through posting and interacting with an online website (Bailey et al.,

2015; Irwin, 2015; Kasket, 2012) in which the bereaved person honors the deceased relative and may

share photographs, music and other memories in honor of the deceased, is therapeutic (i.e., associated

with reduction in intensity of symptoms of PGD). However, we hypothesize that the relationship

between amount of time spent visiting the LMH and reduction of PGD symptom severity will be

curvilinear; that is, too little time spent will not provide much comfort and too much time may foster a

lack of investment in reengaging with the living. Furthermore, around 1 out of 10 bereaved could

develop PGD and are at heightened risk of STBs, and we hypothesize that the linguistic and behavioral

data collected from LMH can be used to predict their future suicide attempt risk.

Primary aim: The main aim is to evaluate and determine how usage of the LMH relates to bereavement

adjustment.

Hypothesis: We hypothesize that the relationship between time spent visiting the LMH and PG-13/BCS

scores will be curvilinear such that there is a middle-range that is most therapeutic and either extremes

of too little time visiting the LMH or too much time visiting the LMH will not be associated with lower

PG-13 scores.

Secondary aim: The pilot study will help researchers to identify the unexpected risks of using the Living

Memory Home and address them for the future deployment.

Tertiary aim: We aim to develop and test an automated suicide attempt risk detection Machine Learning

model using the linguistic and behavioral data collected from LMH.

Hypothesis: Linguistic and behavioral data from these activities will support natural language

processing for classifying suicide attempt risk and the development of machine learning models to

predict suicide attempt risk among bereaved online users.

Research Type:

Weill Cornell Medicine Center for Research on End-of-Life Care 525 E 68th St, Box 39,
1414 Baker Pavilion
New York, NY 10065