Predicting Caller Type From a Mental Health and Well-Being Helpline: Analysis of Call Log Data

Digital transformation of mental health services

This paper makes a case for digital mental health and provides insights into how digital technologies can enhance (but not replace) existing mental health services. We describe digital mental health by presenting a suite of digital technologies (from digital interventions to the application of artificial intelligence). We discuss the benefits of digital mental health, for example, a digital intervention can be an accessible stepping-stone to receiving support. The paper does, however, present less-discussed benefits with new concepts such as 'poly-digital', where many different apps/features (e.g. a sleep app, mood logging app and a mindfulness app, etc.) can each address different factors of wellbeing, perhaps resulting in an aggregation of marginal gains. Another benefit is that digital mental health offers the ability to collect high-resolution real-world client data and provide client monitoring outside of therapy sessions. These data can be collected using digital phenotyping and ecological momentary assessment techniques (i.e. repeated mood or scale measures via an app). This allows digital mental health tools and real-world data to inform therapists and enrich face-to-face sessions. This can be referred to as blended care/adjunctive therapy where service users can engage in 'channel switching' between digital and non-digital (face-to-face) interventions providing a more integrated service. This digital integration can be referred to as a kind of 'digital glue' that helps join up the in-person sessions with the real world. The paper presents the challenges, for example, the majority of mental health apps are maybe of inadequate quality and there is a lack of user retention. There are also ethical challenges, for example, with the perceived 'over-promotion' of screen-time and the perceived reduction in care when replacing humans with 'computers', and the trap of 'technological solutionism' whereby technology can be naively presumed to solve all problems. Finally, we argue for the need to take an evidence-based, systems thinking and co-production approach in the form of stakeholder-centred design when developing digital mental health services based on technologies. The main contribution of this paper is the integration of ideas from many different disciplines as well as the framework for blended care using 'channel switching' to showcase how digital data and technology can enrich physical services. Another contribution is the emergence of 'poly-digital' and a discussion on the challenges of digital mental health, specifically 'digital ethics'.

Forecasting call and chat volumes at online helplines for mental health

BACKGROUND

Each year, many help seekers in need contact health helplines for mental support. It is crucial that they receive support immediately, and that waiting times are minimal. In order to minimize delay, helplines must have adequate staffing levels, especially during peak hours. This has raised the need for means to predict the call and chat volumes ahead of time accurately. Motivated by this, in this paper, we analyze real-life data to develop models for accurately forecasting call volumes, for both phone and chat conversations for online mental health support.

METHODS

This research was conducted on real call and chat data (adequately anonymized) provided by 113 Suicide Prevention (Over ons | 113 Zelfmoordpreventie) (throughout referred to as '113'), the online helpline for suicide prevention in the Netherlands. Chat and phone call data were analyzed to better understand the important factors that influence the call arrival process. These factors were then used as input to several Machine Learning (ML) models to forecast the number of call and chat arrivals. Next to that, senior counselors of the helpline completed a web-based questionnaire after each shift to assess their perception of the workload.

RESULTS

This study has led to several remarkable and key insights. First, the most important factors that determine the call volumes for the helpline are the trend, and weekly and daily cyclic patterns (cycles), while monthly and yearly cycles were found to be non-significant predictors for the number of phone and chat conversations. Second, media events that were included in this study only have limited-and only short-term-impact on the call volumes. Third, so-called (S)ARIMA models are shown to lead to the most accurate prediction in the case of short-term forecasting, while simple linear models work best for long-term forecasting. Fourth, questionnaires filled in by senior counselors show that the experienced workload is mainly correlated to the number of chat conversations compared to phone calls.

CONCLUSION

(S)ARIMA models can best be used to forecast the number of daily chats and phone calls with a MAPE of less than 10 in short-term forecasting. These models perform better than other models showing that the number of arrivals depends on historical data. These forecasts can be used as support for planning the number of counselors needed. Furthermore, the questionnaire data show that the workload experienced by senior counselors is more dependent on the number of chat arrivals and less on the number of available agents, showing the value of insight into the arrival process of conversations.

Machine-learning based routing of callers in an Israeli mental health hotline

Background

Mental health contact centers (also known as Hotlines) offer crisis intervention and counselling by phone calls and online chats. These mental health helplines have shown great success in improving the mental state of the callers, and are increasingly becoming popular in Israel and worldwide. Unfortunately, our knowledge about how to conduct successful routing of callers to counselling agents has been limited due to lack of large-scale data with labeled outcomes of the interactions. To date, many of these contact centers are overwhelmed by chat requests and operate in a simple first-come-first-serve (FCFS) scheduling policy which, combined, may lead to many callers receiving suboptimal counselling or abandoning the service before being treated. In this work our goal is to improve the efficiency of mental health contact centers by using a novel machine-learning based routing policy.

Methods

We present a large-scale machine learning-based analysis of real-world data from the online contact center of ERAN, the Israeli Association for Emotional First Aid. The data includes over 35,000 conversations over a 2-years period. Based on this analysis, we present a novel call routing method, that integrates advanced AI-techniques including the Monte Carlo tree search algorithm. We conducted an experiment that included various realistic simulations of incoming calls to contact centers, based on data from ERAN. We divided the simulations into two common settings: standard call flow and heavy call flow. In order to establish a baseline, we compared our proposed solution to two baseline methods: (1) The FCFS method; and (2) a greedy solution based on machine learning predictions. Our comparison focuses on two metrics - the number of calls served and the average feedback of the callers (i.e., quality of the chats).

Results

In the preliminary analysis, we identify indicative features that significantly contribute to the effectiveness of a conversation and demonstrate high accuracy in predicting the expected duration and the callers’ feedback. In the routing methods evaluation, we find that in heavy call flow settings, our proposed method significantly outperforms the other methods in both the quantity of served calls and average feedback. Most notably, we find that in the heavy call flow settings, our method improves the average feedback by 24% compared to FCFS and by 4% compared to the greedy solution. Regarding the standard-flow setting, we find that our proposed method significantly outperforms the FCFS method in the callers’ average feedback with a 12% improvement. However, in this setting, we did not find a significant difference between all methods in the quantity of served-calls and no significant difference was found between our proposed method and the greedy solution.

Conclusion

The proposed routing policy has the potential to significantly improve the performance of mental health contact centers, especially in peak hours. Leveraging artificial intelligence techniques, such as machine learning algorithms, combined with real-world data can bring about a significant and necessary leap forward in the way mental health hotlines operate and consequently reduce the burden of mental illnesses on health systems. However, implementation and evaluation in an operational contact center is necessary in order to verify that the results replicate in practice.

Artificial Intelligence for Modeling Real Estate Price Using Call Detail Records and Hybrid Machine Learning Approach

Advancement of accurate models for predicting real estate price is of utmost importance for urban development and several critical economic functions. Due to the significant uncertainties and dynamic variables, modeling real estate has been studied as complex systems. In this study, a novel machine learning method is proposed to tackle real estate modeling complexity. Call detail records (CDR) provides excellent opportunities for in-depth investigation of the mobility characterization. This study explores the CDR potential for predicting the real estate price with the aid of artificial intelligence (AI). Several essential mobility entropy factors, including dweller entropy, dweller gyration, workers’ entropy, worker gyration, dwellers’ work distance, and workers’ home distance, are used as input variables. The prediction model is developed using the machine learning method of multi-layered perceptron (MLP) trained with the evolutionary algorithm of particle swarm optimization (PSO). Model performance is evaluated using mean square error (MSE), sustainability index (SI), and Willmott’s index (WI). The proposed model showed promising results revealing that the workers’ entropy and the dwellers’ work distances directly influence the real estate price. However, the dweller gyration, dweller entropy, workers’ gyration, and the workers’ home had a minimum effect on the price. Furthermore, it is shown that the flow of activities and entropy of mobility are often associated with the regions with lower real estate prices.

Engagement With Crisis Text Line Among Subgroups of Users Who Reported Suicidality

OBJECTIVE

Despite the interest in technology-enhanced preventive interventions for suicidality, there is minimal peer-reviewed research on conversations of text message hotlines. In this large-scale study, the authors explored distinct classes of users of the Crisis Text Line who reported suicidality. Classes were based on texters' presenting psychosocial issues and were explored across frequency of hotline use and conversation number.

METHODS

Data included 153,514 conversations from 122,909 individuals collected by the Crisis Text Line (2013-2017). Analyses were restricted to conversations from users who mentioned current or previous suicidality and excluded texters who discussed a third party. The authors used latent class analysis to identify distinct classes of texters on the basis of crisis counselor-assigned issue tags and across subgroups reflecting whether the conversation was the first or last contact with the hotline and frequency of use (one-time, two-time, or ≥3-time texters).

RESULTS

Three classes emerged in all subsamples. The largest class, "lower distress," had the lowest prevalence of all issues. The second largest class, "anxious distress," had the highest prevalence of anxiety or stress and elevated depression. The smallest class, "relational distress," had the highest prevalence of depression and self-harm and higher probability of endorsing relational indicators.

CONCLUSIONS

Psychological and relational issues mostly distinguished the three classes. Despite differing frequency of hotline usage, most texters who reported suicidal ideation and behaviors endorsed similar issues, and these issues did not seem to vary across conversations. Yet there appeared to be distinct subgroups of texters with different presenting issues, which may inform how crisis counselors tailor strategies for both low- and high-volume texters.

Data analytics of call log data to identify caller behaviour patterns from a mental health and well-being helpline

This work presents an analysis of 3.5 million calls made to a mental health and well-being helpline, seeking to answer the question, what different groups of callers can be characterised by specific usage patterns? Calls were extracted from a telephony informatics system. Each call was logged with a date, time, duration and a unique identifier allowing for repeat caller analysis. We utilized data mining techniques to reveal new insights into help-seeking behaviours. Analysis was carried out using unsupervised machine learning (K-means clustering) to discover the types of callers, and Fourier transform was used to ascertain periodicity in calls. Callers can be clustered into five or six caller groups that offer a meaningful interpretation. Cluster groups are stable and re-emerge regardless of which year is considered. The volume of calls exhibits strong repetitive intra-day and intra-week patterns. Intra-month repetitions are absent. This work provides new data-driven findings to model the type and behaviour of callers seeking mental health support. It offers insights for computer-mediated and telephony-based helpline management.