Don't judge a book or health app by its cover: User ratings and downloads are not linked to quality.

OBJECTIVE: To analyse the relationship between health app quality with user ratings and the number of downloads of corresponding health apps. MATERIALS AND METHODS: Utilising a dataset of 881 Android-based health apps, assessed via the 300-point objective Organisation for the Review of Care and Health Applications (ORCHA) assessment tool, we explored whether subjective user-level indicators of quality (user ratings and downloads) correlate with objective quality scores in the domains of user experience, data privacy and professional/clinical assurance. For this purpose, we applied spearman correlation and multiple linear regression models. RESULTS: For user experience, professional/clinical assurance and data privacy scores, all models had very low adjusted R squared values (< .02). Suggesting that there is no meaningful link between subjective user ratings or the number of health app downloads and objective quality measures. Spearman correlations suggested that prior downloads only had a very weak positive correlation with user experience scores (Spearman = .084, p = .012) and data privacy scores (Spearman = .088, p = .009). There was a very weak negative correlation between downloads and professional/clinical assurance score (Spearman = -.081, p = .016). Additionally, user ratings demonstrated a very weak correlation with no statistically significant correlations observed between user ratings and the scores (all p > 0.05). For ORCHA scores multiple linear regression had adjusted R-squared = -.002. CONCLUSION: This study highlights that widely available proxies which users may perceive to signify the quality of health apps, namely user ratings and downloads, are inaccurate predictors for estimating quality. This indicates the need for wider use of quality assurance methodologies which can accurately determine the quality, safety, and compliance of health apps. Findings suggest more should be done to enable users to recognise high-quality health apps, including digital health literacy training and the provision of nationally endorsed "libraries".

Quality of Digital Health Interventions Across Different Health Care Domains: Secondary Data Analysis Study.

BACKGROUND: There are more than 350,000 digital health interventions (DHIs) in the app stores. To ensure that they are effective and safe to use, they should be assessed for compliance with best practice standards. OBJECTIVE: The objective of this paper was to examine and compare the compliance of DHIs with best practice standards and adherence to user experience (UX), professional and clinical assurance (PCA), and data privacy (DP). METHODS: We collected assessment data from 1574 DHIs using the Organisation for the Review of Care and Health Apps Baseline Review (OBR) assessment tool. As part of the assessment, each DHI received a score out of 100 for each of the abovementioned areas (ie, UX, PCA, and DP). These 3 OBR scores are combined to make up the overall ORCHA score (a proxy for quality). Inferential statistics, probability distributions, Kruskal-Wallis, Wilcoxon rank sum test, Cliff delta, and Dunn tests were used to conduct the data analysis. RESULTS: We found that 57.3% (902/1574) of the DHIs had an Organisation for the Review of Care and Health Apps (ORCHA) score below the threshold of 65. The overall median OBR score (ORCHA score) for all DHIs was 61.5 (IQR 51.0-73.0) out of 100. A total of 46.2% (12/26) of DHI's health care domains had a median equal to or above the ORCHA threshold score of 65. For the 3 assessment areas (UX, DP, and PCA), DHIs scored the highest for the UX assessment 75.2 (IQR 70.0-79.6), followed by DP 65.1 (IQR 55.0-73.4) and PCA 49.6 (IQR 31.9-76.1). UX scores had the least variance (SD 13.9), while PCA scores had the most (SD 24.8). Respiratory and urology DHIs were consistently highly ranked in the National Institute for Health and Care Excellence Evidence Standards Framework tiers B and C based on their ORCHA score. CONCLUSIONS: There is a high level of variability in the ORCHA scores of DHIs across different health care domains. This suggests that there is an urgent need to improve compliance with best practices in some health care areas. Possible explanations for the observed differences might include varied market maturity and commercial interests within the different health care domains. More investment to support the development of higher-quality DHIs in areas such as ophthalmology, allergy, women's health, sexual health, and dental care may be needed.

A Multilingual Digital Mental Health and Well-Being Chatbot (ChatPal): Pre-Post Multicenter Intervention Study.

BACKGROUND: In recent years, advances in technology have led to an influx of mental health apps, in particular the development of mental health and well-being chatbots, which have already shown promise in terms of their efficacy, availability, and accessibility. The ChatPal chatbot was developed to promote positive mental well-being among citizens living in rural areas. ChatPal is a multilingual chatbot, available in English, Scottish Gaelic, Swedish, and Finnish, containing psychoeducational content and exercises such as mindfulness and breathing, mood logging, gratitude, and thought diaries. OBJECTIVE: The primary objective of this study is to evaluate a multilingual mental health and well-being chatbot (ChatPal) to establish if it has an effect on mental well-being. Secondary objectives include investigating the characteristics of individuals that showed improvements in well-being along with those with worsening well-being and applying thematic analysis to user feedback. METHODS: A pre-post intervention study was conducted where participants were recruited to use the intervention (ChatPal) for a 12-week period. Recruitment took place across 5 regions: Northern Ireland, Scotland, the Republic of Ireland, Sweden, and Finland. Outcome measures included the Short Warwick-Edinburgh Mental Well-Being Scale, the World Health Organization-Five Well-Being Index, and the Satisfaction with Life Scale, which were evaluated at baseline, midpoint, and end point. Written feedback was collected from participants and subjected to qualitative analysis to identify themes. RESULTS: A total of 348 people were recruited to the study (n=254, 73% female; n=94, 27% male) aged between 18 and 73 (mean 30) years. The well-being scores of participants improved from baseline to midpoint and from baseline to end point; however, improvement in scores was not statistically significant on the Short Warwick-Edinburgh Mental Well-Being Scale (P=.42), the World Health Organization-Five Well-Being Index (P=.52), or the Satisfaction With Life Scale (P=.81). Individuals that had improved well-being scores (n=16) interacted more with the chatbot and were significantly younger compared to those whose well-being declined over the study (P=.03). Three themes were identified from user feedback, including "positive experiences," "mixed or neutral experiences," and "negative experiences." Positive experiences included enjoying exercises provided by the chatbot, while most of the mixed, neutral, or negative experiences mentioned liking the chatbot overall, but there were some barriers, such as technical or performance errors, that needed to be overcome. CONCLUSIONS: Marginal improvements in mental well-being were seen in those who used ChatPal, albeit nonsignificant. We propose that the chatbot could be used along with other service offerings to complement different digital or face-to-face services, although further research should be carried out to confirm the effectiveness of this approach. Nonetheless, this paper highlights the need for blended service offerings in mental health care.

A Mental Health and Well-Being Chatbot: User Event Log Analysis.

BACKGROUND: Conversational user interfaces, or chatbots, are becoming more popular in the realm of digital health and well-being. While many studies focus on measuring the cause or effect of a digital intervention on people's health and well-being (outcomes), there is a need to understand how users really engage and use a digital intervention in the real world. OBJECTIVE: In this study, we examine the user logs of a mental well-being chatbot called ChatPal, which is based on the concept of positive psychology. The aim of this research is to analyze the log data from the chatbot to provide insight into usage patterns, the different types of users using clustering, and associations between the usage of the app's features. METHODS: Log data from ChatPal was analyzed to explore usage. A number of user characteristics including user tenure, unique days, mood logs recorded, conversations accessed, and total number of interactions were used with k-means clustering to identify user archetypes. Association rule mining was used to explore links between conversations. RESULTS: ChatPal log data revealed 579 individuals older than 18 years used the app with most users being female (n=387, 67%). User interactions peaked around breakfast, lunchtime, and early evening. Clustering revealed 3 groups including "abandoning users" (n=473), "sporadic users" (n=93), and "frequent transient users" (n=13). Each cluster had distinct usage characteristics, and the features were significantly different (P<.001) across each group. While all conversations within the chatbot were accessed at least once by users, the "treat yourself like a friend" conversation was the most popular, which was accessed by 29% (n=168) of users. However, only 11.7% (n=68) of users repeated this exercise more than once. Analysis of transitions between conversations revealed strong links between "treat yourself like a friend," "soothing touch," and "thoughts diary" among others. Association rule mining confirmed these 3 conversations as having the strongest linkages and suggested other associations between the co-use of chatbot features. CONCLUSIONS: This study has provided insight into the types of people using the ChatPal chatbot, patterns of use, and associations between the usage of the app's features, which can be used to further develop the app by considering the features most accessed by users.

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'.

System Usability Scale Benchmarking for Digital Health Apps: Meta-analysis.

BACKGROUND: The System Usability Scale (SUS) is a widely used scale that has been used to quantify the usability of many software and hardware products. However, the SUS was not specifically designed to evaluate mobile apps, or in particular digital health apps (DHAs). OBJECTIVE: The aim of this study was to examine whether the widely used SUS distribution for benchmarking (mean 68, SD 12.5) can be used to reliably assess the usability of DHAs. METHODS: A search of the literature was performed using the ACM Digital Library, IEEE Xplore, CORE, PubMed, and Google Scholar databases to identify SUS scores related to the usability of DHAs for meta-analysis. This study included papers that published the SUS scores of the evaluated DHAs from 2011 to 2021 to get a 10-year representation. In total, 117 SUS scores for 114 DHAs were identified. R Studio and the R programming language were used to model the DHA SUS distribution, with a 1-sample, 2-tailed t test used to compare this distribution with the standard SUS distribution. RESULTS: The mean SUS score when all the collected apps were included was 76.64 (SD 15.12); however, this distribution exhibited asymmetrical skewness (-0.52) and was not normally distributed according to Shapiro-Wilk test (P=.002). The mean SUS score for "physical activity" apps was 83.28 (SD 12.39) and drove the skewness. Hence, the mean SUS score for all collected apps excluding "physical activity" apps was 68.05 (SD 14.05). A 1-sample, 2-tailed t test indicated that this health app SUS distribution was not statistically significantly different from the standard SUS distribution (P=.98). CONCLUSIONS: This study concludes that the SUS and the widely accepted benchmark of a mean SUS score of 68 (SD 12.5) are suitable for evaluating the usability of DHAs. We speculate as to why physical activity apps received higher SUS scores than expected. A template for reporting mean SUS scores to facilitate meta-analysis is proposed, together with future work that could be done to further examine the SUS benchmark scores for DHAs.

The impact of the COVID-19 pandemic on grocery shopper behaviour: Analysis of shopper behaviour change using store transaction data.

The systemic shock of coronavirus (COVID-19) and its impact on the global economy has been unprecedented with grocery shopper behaviour changing dramatically through various stages of the pandemic. COVID-19 has caused unusual market conditions, with significant changes to grocery shopper behaviour that need to be understood to allow for appreciation of shopper behaviour change and retail planning implications during future systemic shocks. The aim of this study was therefore to understand grocery-shopping behaviour during COVID-19. Specific objectives were to investigate changes to grocery sale patterns by basket size, composition and category, as well as during specific time periods of the pandemic. The use of transaction data using a range of market basket indicators (e.g., value, size, product mix), revealed profound changes that indicate the challenge shoppers faced navigating a new 'normal grocery shop' and the pressure on retailers to analyse consumption changes in order to prioritise demand planning. While the use of this data and analysis approach is an important contribution to consumer behaviour research, our focus was on the bigger patterns observed through the data pertaining to changes in shopper behaviour during systemic shocks. A key contribution of this paper is how the use of transaction data from grocery retail provides a nuanced understanding of how grocery shoppers responded leading up to and during the pandemic. For example, we found that grocery shoppers purchased more than just 'daily staples' to stock-up during the pandemic, with increased awareness of health and wellbeing an important aspect.

Upper Limb Position Tracking with a Single Inertial Sensor Using Dead Reckoning Method with Drift Correction Techniques.

Inertial sensors are widely used in human motion monitoring. Orientation and position are the two most widely used measurements for motion monitoring. Tracking with the use of multiple inertial sensors is based on kinematic modelling which achieves a good level of accuracy when biomechanical constraints are applied. More recently, there is growing interest in tracking motion with a single inertial sensor to simplify the measurement system. The dead reckoning method is commonly used for estimating position from inertial sensors. However, significant errors are generated after applying the dead reckoning method because of the presence of sensor offsets and drift. These errors limit the feasibility of monitoring upper limb motion via a single inertial sensing system. In this paper, error correction methods are evaluated to investigate the feasibility of using a single sensor to track the movement of one upper limb segment. These include zero velocity update, wavelet analysis and high-pass filtering. The experiments were carried out using the nine-hole peg test. The results show that zero velocity update is the most effective method to correct the drift from the dead reckoning-based position tracking. If this method is used, then the use of a single inertial sensor to track the movement of a single limb segment is feasible.

Discovering and comparing types of general practitioner practices using geolocational features and prescribing behaviours by means of K-means clustering.

Traditionally General Practitioner (GP) practices have been labelled as being in Rural, Urban or Semi-Rural areas with no statistical method of identifying which practices fall into each category. The main aim of this study is to investigate whether location and other characteristics can provide a tautology to identify different types of GP practice and compare the prescribing behaviours associated with the different practice types. To achieve this monthly open source prescription data were analysed by practice considering location, practice size, population density and deprivation rankings. One year's data was subjected to k-means clustering with the results showing that only two different types of GP practice can be classified that are dependent on location characteristics in Northern Ireland. Traditional labels did not describe the two classifications fully and new classifications of Metropolitan and Non-Metropolitan were used. Whilst prescribing patterns were generally similar, it was found that Metropolitan practices generally had higher prescribing rates than Non-Metropolitan practices. Examining prescribing behaviours in accordance with British National Formulary (BNF) categories (known as chapters) showed that Chapter 4 (Central Nervous System) was responsible for most of the difference in prescribing levels. Within Chapter 4 higher prescribing levels were attributable to Analgesic and Antidepressant prescribing. The clusters were finally examined regarding the level of deprivation experienced in the area in which the practice was located. This showed that the Metropolitan cluster, having higher prescription rates, also had a higher proportion of practices located in highly deprived areas making deprivation a contributing factor.

Digital reminiscence app co-created by people living with dementia and carers: Usability and eye gaze analysis.

BACKGROUND: This research reports on a pilot study that examined the usability of a reminiscence app called 'InspireD' using eye tracking technology. The InspireD app is a bespoke digital intervention aimed at supporting personalized reminiscence for people living with dementia and their carers. The app was developed and refined in two co-creation workshops and subsequently tested in a third workshop using eye tracking technology. INTERVENTION: Eye tracking was used to gain insight into the user's cognition since our previous work showed that the think-aloud protocol can add to cognitive burden for people living with dementia while also making the test more unnatural. RESULTS: Results showed that there were no barriers to using a wearable eye tracker in this setting and participants were able to use the reminiscence app freely. However, some tasks required prompts from the observer when difficulties arose. While prompts are not normally used in usability testing (as some argue the prompting defeats the purpose of testing), we used 'prompt frequency' as a proxy for measuring the intuitiveness of the task. There was a correlation between task completion rates and prompt frequency. Results also showed that people living with dementia had fewer gaze fixations when compared to their carers. Carers had greater fixation and saccadic frequencies when compared to people living with dementia. This perhaps indicates that people living with dementia take more time to scan and consume information on an app. A number of identified usability issues are also discussed in the paper. PATIENT OR PUBLIC CONTRIBUTION: The study presents findings from three workshops which looked at user needs analysis, feedback and an eye tracking usability test combined involving 14 participants, 9 of whom were people living with dementia and the remaining 5 were carers.

Ethical Issues in Democratizing Digital Phenotypes and Machine Learning in the Next Generation of Digital Health Technologies.

Digital phenotyping is the term given to the capturing and use of user log data from health and wellbeing technologies used in apps and cloud-based services. This paper explores ethical issues in making use of digital phenotype data in the arena of digital health interventions. Products and services based on digital wellbeing technologies typically include mobile device apps as well as browser-based apps to a lesser extent, and can include telephony-based services, text-based chatbots, and voice-activated chatbots. Many of these digital products and services are simultaneously available across many channels in order to maximize availability for users. Digital wellbeing technologies offer useful methods for real-time data capture of the interactions of users with the products and services. It is possible to design what data are recorded, how and where it may be stored, and, crucially, how it can be analyzed to reveal individual or collective usage patterns. The paper also examines digital phenotyping workflows, before enumerating the ethical concerns pertaining to different types of digital phenotype data, highlighting ethical considerations for collection, storage, and use of the data. A case study of a digital health app is used to illustrate the ethical issues. The case study explores the issues from a perspective of data prospecting and subsequent machine learning. The ethical use of machine learning and artificial intelligence on digital phenotype data and the broader issues in democratizing machine learning and artificial intelligence for digital phenotype data are then explored in detail.

An analysis of the impact of suicide prevention messages and memorials on motorway bridges.

Recently, there has been activity at public locations where people have died by suicide, including the erection of suicide prevention messages and memorials (decorations). This research looks at the impact of these decorations and associated media coverage of the decorations on suicidal behaviour at bridges. Incidents (n = 160) of suicidal behaviour on 26 bridges across motorways in England were analysed. Overall, there was no significant difference in the proportion of incidents pre-decoration versus post-decoration (p-value = .55). The incident rates were not significantly different pre- and post-decoration (p = .46). Only one bridge had statistically significantly more incidents post-decoration and media reporting (p = .03). However, following correction for multiple testing there was no significant difference in pre- and post-incident rates at any of the bridges. In total, 58% of bridges had a greater frequency of incidents when decorations were absent; however, this proportion was not statistically significant (p = .41). Further research is required to establish how suicide prevention messages are perceived. There does not appear to be any benefit, but it often generates media coverage which has been shown to increase risk.

Behavior of Callers to a Crisis Helpline Before and During the COVID-19 Pandemic: Quantitative Data Analysis.

BACKGROUND: The World Health Organization declared the outbreak of COVID-19 to be an international pandemic in March 2020. While numbers of new confirmed cases of the disease and death tolls are rising at an alarming rate on a daily basis, there is concern that the pandemic and the measures taken to counteract it could cause an increase in distress among the public. Hence, there could be an increase in need for emotional support within the population, which is complicated further by the reduction of existing face-to-face mental health services as a result of measures taken to limit the spread of the virus. OBJECTIVE: The objective of this study was to determine whether the COVID-19 pandemic has had any influence on the calls made to Samaritans Ireland, a national crisis helpline within the Republic of Ireland. METHODS: This study presents an analysis of calls made to Samaritans Ireland in a four-week period before the first confirmed case of COVID-19 (calls=41,648, callers=3752) and calls made to the service within a four-week period after a restrictive lockdown was imposed by the government of the Republic of Ireland (calls=46,043, callers=3147). Statistical analysis was conducted to explore any differences between the duration of calls in the two periods at a global level and at an hourly level. We performed k-means clustering to determine the types of callers who used the helpline based on their helpline call usage behavior and to assess the impact of the pandemic on the caller type usage patterns. RESULTS: The analysis revealed that calls were of a longer duration in the postlockdown period in comparison with the pre-COVID-19 period. There were changes in the behavior of individuals in the cluster types defined by caller behavior, where some caller types tended to make longer calls to the service in the postlockdown period. There were also changes in caller behavior patterns with regard to the time of day of the call; variations were observed in the duration of calls at particular times of day, where average call durations increased in the early hours of the morning. CONCLUSIONS: The results of this study highlight the impact of COVID-19 on a national crisis helpline service. Statistical differences were observed in caller behavior between the prelockdown and active lockdown periods. The findings suggest that service users relied on crisis helpline services more during the lockdown period due to an increased sense of isolation, worsening of underlying mental illness due to the pandemic, and reduction or overall removal of access to other support resources. Practical implications and limitations are discussed.

Reliability of Supervised Machine Learning Using Synthetic Data in Health Care: Model to Preserve Privacy for Data Sharing.

BACKGROUND: The exploitation of synthetic data in health care is at an early stage. Synthetic data could unlock the potential within health care datasets that are too sensitive for release. Several synthetic data generators have been developed to date; however, studies evaluating their efficacy and generalizability are scarce. OBJECTIVE: This work sets out to understand the difference in performance of supervised machine learning models trained on synthetic data compared with those trained on real data. METHODS: A total of 19 open health datasets were selected for experimental work. Synthetic data were generated using three synthetic data generators that apply classification and regression trees, parametric, and Bayesian network approaches. Real and synthetic data were used (separately) to train five supervised machine learning models: stochastic gradient descent, decision tree, k-nearest neighbors, random forest, and support vector machine. Models were tested only on real data to determine whether a model developed by training on synthetic data can used to accurately classify new, real examples. The impact of statistical disclosure control on model performance was also assessed. RESULTS: A total of 92% of models trained on synthetic data have lower accuracy than those trained on real data. Tree-based models trained on synthetic data have deviations in accuracy from models trained on real data of 0.177 (18%) to 0.193 (19%), while other models have lower deviations of 0.058 (6%) to 0.072 (7%). The winning classifier when trained and tested on real data versus models trained on synthetic data and tested on real data is the same in 26% (5/19) of cases for classification and regression tree and parametric synthetic data and in 21% (4/19) of cases for Bayesian network-generated synthetic data. Tree-based models perform best with real data and are the winning classifier in 95% (18/19) of cases. This is not the case for models trained on synthetic data. When tree-based models are not considered, the winning classifier for real and synthetic data is matched in 74% (14/19), 53% (10/19), and 68% (13/19) of cases for classification and regression tree, parametric, and Bayesian network synthetic data, respectively. Statistical disclosure control methods did not have a notable impact on data utility. CONCLUSIONS: The results of this study are promising with small decreases in accuracy observed in models trained with synthetic data compared with models trained with real data, where both are tested on real data. Such deviations are expected and manageable. Tree-based classifiers have some sensitivity to synthetic data, and the underlying cause requires further investigation. This study highlights the potential of synthetic data and the need for further evaluation of their robustness. Synthetic data must ensure individual privacy and data utility are preserved in order to instill confidence in health care departments when using such data to inform policy decision-making.

Ecological Momentary Assessment Within a Digital Health Intervention for Reminiscence in Persons With Dementia and Caregivers: User Engagement Study.

BACKGROUND: User-interaction event logs provide rich and large data sets that can provide valuable insights into how people engage with technology. Approaches such as ecological momentary assessment (EMA) can be used to gather accurate real-time data in an individual's natural environment by asking questions at any given instant. OBJECTIVE: The purpose of this study was to evaluate user engagement and responses to EMA questions using InspireD, an app used for reminiscence by persons with dementia and their caregivers. Research findings can be used to inform EMA use within digital health interventions. METHODS: A feasibility trial was conducted in which participants (n=56) used the InspireD app over a 12-week period. Participants were a mean age of 73 (SD 13) and were either persons with dementia (n=28) or their caregivers (n=28). Questions, which they could either answer or choose to dismiss, were presented to participants at various instants after reminiscence with personal or generic photos, videos, and music. Presentation and dismissal rates for questions were compared by hour of the day and by trial week to investigate user engagement. RESULTS: Overall engagement was high, with 69.1% of questions answered when presented. Questions that were presented in the evening had the lowest dismissal rate; the dismissal rate for questions presented at 9 PM was significantly lower than the dismissal rate for questions presented at 11 AM (9 PM: 10%; 11 AM: 50%; χ21=21.4, P<.001). Questions asked following reminiscence with personal media, especially those asked after personal photos, were less likely to be answered compared to those asked after other media. In contrast, questions asked after the user had listened to generic media, in particular those asked after generic music, were much more likely to be answered. CONCLUSIONS: The main limitation of our study was the lack of generalizability of results to a larger population given the quasi-experimental design and older demographic where half of participants were persons with dementia; however, this study shows that older people are willing to participate and engage in EMA. Based on this study, we propose a series of recommendations for app design to increase user engagement with EMA. These include presenting questions no more than once per day, after 8 PM in the evening, and only if the user is not trying to complete a task within the app.

Why do people call crisis helplines? Identifying taxonomies of presenting reasons and discovering associations between these reasons.

The objective of this study is to identify the most common reasons for contacting a crisis helpline through analysing a large call log data set. Two taxonomies were identified within the call log data from a Northern Ireland telephone crisis helpline (Lifeline), categorising the cited reason for each call. One taxonomy categorised the reasons at a fine granular level; the other taxonomy used the relatively coarser International Classification of Diseases-10. Exploratory data analytic techniques were applied to discover insights into why individuals contact crisis helplines. Risk ratings of calls were also compared to assess the associations between presenting issue and of risk of suicide as assessed. Reasons for contacting the service were assessed across geolocations. Association rule mining was used to identify associations between the presenting reasons for client's calls. Results demonstrate that both taxonomies show that calls with reasons relating to suicide are the most common reasons for contacting Lifeline and were a prominent feature of the discovered association rules. There were significant differences between reasons in both taxonomies concerning risk ratings. Reasons for calling helplines that are associated with higher risk ratings include those calling with a personality disorder, mental disorders, delusional disorders and drugs (legal). In conclusion, employing two differing taxonomy approaches to analyse call log data reveals the prevalence of main presenting reasons for contacting a crisis helpline. The association rule mining using each taxonomy provided insights into the associations between presenting reasons. Practical and research applications are discussed.

Economic costs and health-related quality of life associated with individual specific reminiscence: Results from the InspireD Feasibility Study.

Recent studies have focused on the use of technology to support reminiscence but there remains a dearth of research on the health costs and benefits associated with this intervention. The aim of this study was to estimate costs and quality of life associated with a home based, individual specific reminiscence intervention, facilitated by an iPad app for people living with dementia and their family carers, with a view to informing a future cost-effectiveness analysis. Use of community health and social care services, hospital services, prescribed medication and informal caregiving was assessed using an adapted version of the Client and Socio-Demographic Service Receipt Inventory (CSRI) at baseline and 3-month follow-up. Quality of life was assessed at baseline, 6-week and 3-month follow-up using the EQ5D, DEMQOL and DEMQOL proxy instruments. Results showed that average health and social care costs were £29,728 per person at baseline (T0) and £33,436 after 3 months (T2). Higher T2 costs were largely accounted for by higher informal caregiving costs. There was an overall increase in health-related quality of life over the duration of the intervention, although there were notable differences in index scores generated by the EQ5D (0.649, 0.652 and 0.719) and DEMQOL instruments (0.845, 0.968 and 0.901). The study concluded that a full cost-effectiveness analysis could incorporate a similar range of cost-categories with minor amendments to the CSRI to improve the accuracy of cost estimation. Furthermore, a larger sample size, randomisation and longer follow-up period are required to allow potential effects of the intervention to be realised and differences between intervention and control groups to be accurately detected.

Evaluating User Engagement with a Reminiscence App Using Cross-Comparative Analysis of User Event Logs and Qualitative Data.

The aim of this study was to evaluate the usage of a reminiscence app by people living with dementia and their family carers, by comparing event log data generated from app usage alongside the qualitative experience of the process. A cross-comparative analysis of electronic event logging data with qualitative interview data was conducted. Electronic event logging data were obtained for 28 participating dyads (n = 56) and the interview sample comprised 14 people living with dementia and 16 family carers (n = 30). A thematic analysis framework was used in the analysis of interview transcripts and the identification of recurrent themes. The cross-comparison of electronic event log data and qualitative data revealed 25 out of 28 dyads regularly engaged with a reminiscence app, with the analysis of usage patterns revealing four clusters classifying different levels of user engagement. The cross-comparison of data revealed that the nature of the relationship was a significant factor in ongoing user engagement. The comparative analysis of the electronic event logs as "ground truth" in combination with the qualitative lived experience can provide a deeper understanding on the usage of a reminiscence app for those living with dementia and their family carers. This work not only shows the benefits of using automated event log data mining but also shows its clear limitations without using complementary qualitative data analysis. As such, this work also provides key insights into using mixed methods for evaluating human-computer interaction technologies.

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.