Abstract
Study objectives: Crisis lines are vital to suicide prevention efforts and provide immediate emotional support to those struggling with loneliness and distress. A knowledge of call volume and caller behaviour patterns is of value in service and operational planning, and also help us under-stand how these lines may reflect, and impact upon, suicidal behaviour and helpseeking.
Methods and materials: This study analyses call data provided by Samaritans UK. Initially, the volume of ongoing calls were calculated with the frequency of 48 samples per day, by considering both number of call arrivals and duration of calls in each half an hour. The daily signals are decomposed using wavelet transform to obtain the approximation (trend-line) and coefficients of the signal.
Results: The preliminary results of this study represent the annual spectrogram of the wavelet approximation for 2015-2017. The spectrograms show that the Samaritans receive the highest volume of calls around 10pm to 2.30am, followed by a substantial reduction of call volume between 5am to 9am, particularly during the final days of the year, on a daily basis. Observation of weekly cycles indicate that the highest call volume throughout the week occurs between Sunday and Tuesday, with the lowest call volume between Thursday and Saturday. Accordingly, daily and weekly repetitive patterns are clearly detectable; however, there was no definite annual cyclic behaviour observed in this 3-year analysis.
Conclusion: Wavelet analysis of call volume can be applied to crisis line caller behaviour and provides insight to inform service planning. The caller behaviour patterns identified may illustrate the times at which people experience crisis.
Methods and materials: This study analyses call data provided by Samaritans UK. Initially, the volume of ongoing calls were calculated with the frequency of 48 samples per day, by considering both number of call arrivals and duration of calls in each half an hour. The daily signals are decomposed using wavelet transform to obtain the approximation (trend-line) and coefficients of the signal.
Results: The preliminary results of this study represent the annual spectrogram of the wavelet approximation for 2015-2017. The spectrograms show that the Samaritans receive the highest volume of calls around 10pm to 2.30am, followed by a substantial reduction of call volume between 5am to 9am, particularly during the final days of the year, on a daily basis. Observation of weekly cycles indicate that the highest call volume throughout the week occurs between Sunday and Tuesday, with the lowest call volume between Thursday and Saturday. Accordingly, daily and weekly repetitive patterns are clearly detectable; however, there was no definite annual cyclic behaviour observed in this 3-year analysis.
Conclusion: Wavelet analysis of call volume can be applied to crisis line caller behaviour and provides insight to inform service planning. The caller behaviour patterns identified may illustrate the times at which people experience crisis.
Original language | English |
---|---|
Publication status | Published (in print/issue) - 21 Oct 2019 |
Event | 30th World Congress of the International Association for Suicide Prevention: Breaking Down Walls & Building Bridges - Millenium Forum, Derry.Londonderry, United Kingdom Duration: 17 Sept 2019 → 21 Sept 2019 Conference number: 30 https://www.iasp2019.com |
Conference
Conference | 30th World Congress of the International Association for Suicide Prevention |
---|---|
Abbreviated title | IASP |
Country/Territory | United Kingdom |
City | Derry.Londonderry |
Period | 17/09/19 → 21/09/19 |
Internet address |
Keywords
- helplines
- Fourier
- wavelet analysis
- Log analysis
- Calls
- call centre
- Mental Health