Confirmed cases, fatalities, hospitalizations and weekly infection growth rate have become the default metrics for understanding the severity of the Covid-19 pandemic in Singapore. However, a symptom impacting most Singaporeans, regardless of their vaccination status, has proven to be harder to quantify: mental health decline.

A post-mortem study conducted by Wu et al concluded that the pandemic had imposed a significant mental health burden on society. This has led to higher prevalence of depression, anxiety, distress and insomnia amongst the population. Closer to home, Samaritans of Singapore (SOS) reported a 30-35% increase in calls to their hotline during the circuit breaker period.

However, this increase in demand is not matched with analogous increase in supply of mental health support avenues. From an outside view in, there seems to be an easy fix to the predicament: hire more providers. Workforce recruitment and retention had been an outstanding issue in the field of behavioral health, with low compensation and limited advancement prospects cited as the main impediments. However, even under the assumption that there are enough hires, a provider would have to undergo sufficient training before they are certified to provide the required services. In the case of SOS, every aspiring provider would have to undergo a mandatory nine month foundation training, then evaluated on a buddy system (tagged with a senior) until he/she is given the greenlight to go solo.

In preparation for the influx of demand from the pandemic, SOS started CareText, a text messaging-based service, in mid-July 2020, three months prior to its planned launch to alleviate the mental health needs in Singapore. While the mode of delivery is not perfect, CareText provide much needed relief for the overtaxed volunteers by allowing SOS to scale. Conventional phone room services would require a 1:1 attention from the volunteers, while CareText provide flexibility by allowing a volunteer to assist up to 3 clients concurrently. Further, the evaluation process would be less resource intensive, freeing up valuable senior expertise to handle more urgent matters.

From a theorist perspective, CareText would seemingly be replicable by an emotional chatbot, if given enough training data. We endeavor to learn more about the current process of CareText and seek to implement technology to assist the volunteers along their process.