An Analysis of Caller Behaviour to a Crisis Helpline during the COVID-19 Pandemic

Robin Turkington, Maurice Mulvenna, RR Bond, Courtney Potts, Edel Ennis, Siobhan O'Neill, Ciaran Moore, Louise Hamra, Jacqui Morrissey, Mette Isaksen, Elizabeth Scowcroft

Research output: Contribution to conferencePosterpeer-review

Abstract

Background
There is growing concern of the impact that the COVID-19 pandemic may have on population mental health. Data indicates that levels of distress have increased due to the pandemic. People living with mental illness may be impacted most as face-to-face support groups have been disrupted due to the lockdown restrictions, leading to a reliance on remote services such as crisis helplines. The objective was to determine whether telephony data could reveal the impact of the COVID-19 pandemic on callers to Samaritans Ireland.

Methods
This study presents an analysis of calls made to Samaritans Ireland within four periods lasting four weeks each; one period before the first confirmed case in Ireland (Pre-COVID-19; 3rd Feb - 1st Mar) and three other periods after the introduction of the lockdown restrictions; a first Active COVID-19 period (30th Mar - 26th April), a second Active COVID-19 period (1st Jun - 28th Jun) and a third Active COVID-19 period (3rd Aug - 30th Aug). Statistical analysis was conducted to explore any differences in duration of calls across all four periods. Clustering was performed to determine caller types that use the helpline based on their usage patterns and whether this changed across the four periods.

Results
The incidence of longer calls increased dramatically in the first Active COVID-19 period from the Pre-COVID-19 period, reflecting higher levels of support being sought. This trend towards longer calls decreased across the later Active COVID-19 periods towards the Pre-COVID-19 period norm . Five caller types were discovered and each showed differences in the time of day in which they made calls and their distribution of call duration across all periods. Significant differences were noticed between the Pre-COVID-19 period and the first Active COVID-19 period. The behaviours observed in the later Active COVID-19 periods trended towards that of a Pre-COVID-19 norm.

Discussion
The findings highlight that telephony data can be used to measure the effect of an external event on society; in this case, the impact of COVID-19 on a national crisis helpline service. The results indicate that the introduction of the lockdown restrictions created a societal impact on the population. The trend back to the Pre-COVID-19 norm in call behaviour may indicate hedonic adaptation amongst callers to the service.

Implications
The findings indicate the viability of using telephony data to provide near real time insights in behaviours beyond the helpline, across broader society impacted by external events.
Original languageEnglish
Publication statusPublished - 17 Nov 2020
EventPublic Health Agency: Annual Scientific Conference -
Duration: 1 Jun 2013 → …

Conference

ConferencePublic Health Agency: Annual Scientific Conference
Period1/06/13 → …

Keywords

  • Mental health
  • Crisis helplines
  • Caller behaviour
  • Machine learning

Fingerprint Dive into the research topics of 'An Analysis of Caller Behaviour to a Crisis Helpline during the COVID-19 Pandemic'. Together they form a unique fingerprint.

Cite this