Artificial Intelligence and Social Work


Social Work and related professions have the potential to experience rapid change and growth in the future as technology advances and the population changes. This is especially true with artificial intelligence.

Artificial Intelligence (AI) describes a range of technologies that allow machines or computers to make decisions that are normally made by human beings.

Emotional Support Technology

Perhaps the first attempt at emotional support using a computer was the ELIZA software created by Joseph Weizenbaum in 1964. Through pattern matching the software was able to respond with empathy statements and open-ended questions to keep the conversation going.

Modern options include XiaoIce (Zhao, et. al., 2018). As the authors describe,

The primary design goal of XiaoIce is to be an AI companion with which users form long-term, emotional connections. Being able to establish such long-term relationships with human users as an open-domain social chatbot distinguishes XiaoIce from not only early social chatbots but also other recently developed conversational AI personal assistants such as Apple Siri, Amazon Alexa, Google Assistant and Microsoft Cortana.

Another example includes Replika, which recently released its source code as open-source. As these technologies get more advanced they may play a more important role in our emotional support options for people who are struggling with loneliness.

Digital Psychotherapy

Digital psychotherapy options include online and electronic therapy options. One example is Electronic CBT for Insomnia (Espie, et. al., 2018) which was a rich-media web application that participants used to receive cognitive behavioral therapy via the internet, and Whiteside et. al. (2014) which studied the program Thrive:

Thrive is similar to programs used in successful trials of Internet-delivered CBT; the Thrive interface is interactive and its curriculum is adaptive to patient input. […] Thrive includes three CBT-based modules that are based on behavioral activation, cognitive restructuring, and social skills training techniques

While these programs are currently not utilizing much artificial intelligence, in the future we may see them adapting to the client’s progress and altering the curriculum in ways that will increase efficacy or completion rates.

As CBT programs become more researched and advance we should see more of these appearing. As Whiteside notes, these programs are significantly cheaper to deliver (using a fully automated or a paraprofessional “coach” model rather than delivering full therapy) and so may represent an increasingly common option for therapists.

Decision-Making Tools

Decision-making tools are potentially the most exciting use of technology and artificial intelligence. An example of where this technology has been helpful is in child protection work in England. (Pegg & McIntyre, 2018)

We may see AI being used in the future to help us integrate the hundreds of variables found in child protection assessments and files to increase our success rates and improve risk assessments. Certainly, we can’t replace humans in this incredibly careful work (just like in suicide risk assessment) but we can use these tools to augment our understanding of child protection and decrease the lag between learning things in research and applying them in practice.


Artificial intelligence has the potential to improve our lives by providing more emotional support to those who are lonely, providing digital psychotherapy and decision-support tools to improve child protection and other social work fields.


Espie, C.A., Kyle, S.D., Williams, C., Ong, J.C., Douglas, N.J., Hames, P., Brown, J.S.L. (2012) JAMA Psychiatry. Retrieved from

Pegg, D. & McIntyre, N. (2018) Child abuse algorithms: from science fiction to cost-cutting reality. Retrieved from

Weizenbaum, Joseph (1976). Computer Power and Human Reason: From Judgment to Calculation. New York: W.H. Freeman and Company. pp. 2, 3, 6, 182, 189. ISBN 0-7167-0464-1.

Whiteside, U., Richards, J., Steinfeld, B., Simon, G., Caka, S., Tachibana, C., Stuckey, S., … Ludman, E. (2014). Online cognitive behavioral therapy for depressed primary care patients: a pilot feasibility project. The Permanente journal18(2), 21-7.

Zhou, L., Gao, J., Li, D. & Heung-Yeung, S. (2018) The Design and Implementation of XiaoIce, an Empathetic Social Chatbot. Journal of Human and Computer Interaction.

Facebooktwittergoogle_plusredditmailby feather

Providing Emotional Support Over Text and Chat

Introduction to Text and ChatRoleplay Transcript

With text and chat services increasingly moving online, emotional support work – the core element of the work of crisis lines is needing to be adapted to work in new formats that require a change in your perspective and technique. On the telephone, there are a number of ways of providing a warm, genuine experience. For instance, your voice tone and pitch communicates a lot, as well as the speed in which you talk, whether you speak over the caller or let them lead, and so on. There is a lot of non-verbal communication that happens on the phone.

In contrast, online all you have is text. So many of the dimensions that are used to promote warmth, communicate empathy and demonstrate caring are simply absent. This makes it more difficult to build rapport with these visitors and be effective.

The elements of active listening, or the active listening process are the same, although of course it seems unusual to call it “listening” since you aren’t using your ears. There is still an effort made to be alert for and respond to communication, however. Some people prefer “emotional support” instead.

Chat and Text Length

Chat and text conversations tend to be longer than telephone conversations; an average telephone call may be 20 minutes while a crisis chat or text conversation will be 45-60 minutes. This is due to the time required for you to send a text, for the visitor to receive it, read it, decide what they’re going to write, and then write back. You may not send a lot of  messages in this 60 minutes, but that doesn’t mean that you aren’t accomplishing a lot – which is reflected in the outcomes, often up to 30% reduction in subjective distress over an hour.

Opening Conversations

In the opening of a text-based conversation, it’s important to be warm and genuine. Your opening message should give your name, because the visitor doesn’t have anything else to go on. You may want or need to identify your organization as well. Finally, you’ll want to ask the visitor what brought them to text in.

An example of an opening message I could use on the ONTX Project is “Welcome to the ONTX Project. My name is Dustin, what’s going on in your life?”

Sometimes a visitor will text in with a lethality statement, something like, “I want to die.” This doesn’t necessarily change your opening, but it doesn’t hurt to acknowledge the suicidal feeling. “Welcome to the ONTX Project. My name is Dustin, it sounds like you’re really struggling. Did you want to tell me what’s been going on?”

Some visitors though, may need a bit of encouragement. If you ask a visitor how they’re feeling, they may reply “idk” (I don’t know) or “bad”, and not elaborate. Other visitors may be much more articulate and be able to explain what’s going on in their life.

If someone says “idk” or “bad”, usually my next move is to ask them what’s on their mind tonight. This is a gentle way of rewording the question that helps them feel more comfortable. Usually at this point they’ll begin talking, but if not my final option is “What were you hoping to get out of texting in tonight?”

I’ve never had a visitor respond with “idk” or other messages after this much encouragement but I would likely empathize with how difficult it’s been for them to text in before ending the conversation and inviting them to try us again when they’re more able to speak.

Because of a 140 character limit, some of these messages may need to be sent as a pair of messages on text.

Exploring the Issue

Exploring the issue that the visitor is texting in about can be challenging. Unlike the helpline, where you may need to take a while to establish rapport, visitors on text tend to jump right to their primary concern rather quickly. They don’t have the luxury of many messages back and forth.

If you’ve used the above Opening the Conversation ideas, you should be well into exploring the issue. This section should proceed just the same way as an offline conversation does, using all elements of the active listening process (open ended questions, paraphrasing and summarizing.)

You may notice that you need to ask more clarifying questions than usual, because with text and a lack of tone it’s easier for things to be misunderstood or misconstrued.

Demonstrating Empathy

In an online environment, you have no voice tone to demonstrate empathy. For this reason it’s important to write out your empathy statements clearly in order to show that you have an idea what the visitor is going through. Clarifying and paraphrasing can help in rapport building as well, by demonstrating that you are paying attention. It’s important to recognize that clarifying, paraphrasing and other open and close-ended questions are not a replacement for pure empathy.

Empathy: You sound really alone.

Clarifying: You just lost your dog?

Paraphrasing: You’ve been having trouble since you lost your pet.

Note the difference, empathy highlights an emotion (alone) while clarifying and paraphrasing primarily on content without regard to an emotional undertone.

Suicide Risk Assessment and Intervention

Suicide risk assessment and intervention is a challenging topic over chat and text. The primary challenges in this environment include the difficulty collecting the amount of information required to perform a competent assessment in 140 characters and the lack of voice tone and body language.

Typically the first question asked on chat and text after confirming suicide thoughts are present is to determine if they’re at imminent risk. This is usually accomplished by asking something like “Have you done anything to kill yourself?” or “Have you taken any steps to end your life tonight?”

Chatters and texters will sometimes text in immediately after an overdose, and will readily reveal their level of danger but not until you ask. Sooner rather than later!

Next, I’ll ask the visitor what’s led them to feeling suicidal. This, when combined with an empathy statements, helps to begin exploring the visitor’s reasons for living or dying. For example, “You must be feeling so overwhelmed. Tell me what’s led you to feeling suicidal?”

After this, I move onto the elements of the DCIB Suicide Risk Assessment tool.

Winding Up Conversations

Because visitors are using their cell phones, they can put their phone in their pocket, and then pull it out without thinking about the time that passes in a few minutes. It’s not uncommon that at the end of your 45-60 minutes, when it comes to winding up, the visitor doesn’t even realize that amount of time has passed. They find themselves feeling better, however, which is great news!

Winding up has to be deliberate, otherwise the visitor is unlikely to wind up in a decent time. Past experience has shown that crisis chats can last 3 hours or longer lacking a proper wind up. In order to initiate a windup, you simply have to give the visitor an opportunity to express anything else on their mind and then let them know that you have to go. For example,

“We’ve been talking for about an hour so we’ll need to wrap our conversation up soon. I’m wondering if there’s anything else on your mind that you haven’t shared yet.”

Or, more succinctly,

“We’re just coming up on 45 minutes of chatting so we’ll need to wind up soon. Was there anything else you wanted to share before we do?”

This cues the visitor that the conversation needs to end and lets them focus on any outstanding issues. For instance, you may be convinced of their safety and they may not be – and by pointing that out by replying “I don’t know what to do to avoid attempting suicide tonight” then you can spend your remaining 15 minutes implementing a comprehensive safety plan for that visitor. In this way, the windup can be a tool for you and the visitor.

Cite this article as: MacDonald, D.K., (2016), "Providing Emotional Support Over Text and Chat," retrieved on July 23, 2019 from
Facebooktwittergoogle_plusredditmailby feather

Beginnings of an Online Crisis Chat Service

On June 29th, the Distress Centres of Dufferin/Wellington, Durham, Peel and Toronto joined together to launch an online crisis chat service. Texting through an unadvertised shortcode based on a service in the US or visiting our website provides access to a trained volunteer responder between 2pm and 2am.

After running the service for a few months, we’ve begun seeing some important elements emerge from the data. They include elements like the length of the chats, the service volume, the elements necessary for high-quality chats and texts,

Chat Length

The average interaction is 45-60 minutes. This is much longer than helpline conversations which tend to be closer to 20-30 minutes for individuals who are not in crisis. This is because of the time required to process what you’re reading and to consider writing your response.

Service Volume

The service operates for approximately 12 hours a day and fields an average of 10-15 text sessions and 5-10 chat sessions a day, but has received as many as 64 chat/text interactions in a week during periods of high volume.

Number of Responders

We are close to 150 responders covering a 12 hour period with double coverage. Because commitments can vary based on the partner Centres we estimated that we would need about 50 responders per Centre in order to meet this base level of service.

Level of Distress

Currently, there is a pre-chat survey for visitors entering the system where they can indicate their level of distress on a 5 point scale:

  • 1 – I’m doing okay
  • 2 – A little upset
  • 3 – Moderately upset
  • 4 – Very upset
  • 5 – Extremely upset

At the end of the chat, they are redirected to a post-chat survey where they have the option to note if their level of distress has changed. Approximately 20% of the chat visitors complete the pre and post chat surveys.

The results of preliminary analysis show that their level of distress drops 1 full point from 3.8 to 2.8, on average after an interaction with a responder.

Chat and Text Issues

The most common issues that chatters report are anxiety, depression, relationship concerns and suicidal thoughts. This is very similar to the helpline, though the rate of suicidal ideation is 2-3x higher (10-15%) on the service than the provincial average on the helpline (approximately 5%.)

This range of issues is similar to the helpline.

Age Range

Right now the service sees 100% of it’s usage between 0-44, no visitors over 45 have visited. The most common age range is 15-24, though I’ve seen some as young as 12 accessing the service. Only 5% of the visitors on the telephone helpline are 15-24, suggesting the service is much more popular for youth.

Crisis Chat / Text Training

The training is approximately 4 hours in length and covers a review of active listening and emotional support, the DCIB Suicide Risk Assessment, how to use the iCarol messaging system and then some time for roleplays.

It’s important to recognize that there is a different skill set required for communication through a text environment versus communication over the telephone. For instance, the lack of voice tone makes it very difficult to recognize when you are de-escalating a visitor.

This required consistent checking in with the visitor to make sure they are actually improving.

Crisis Chat / Text Suicide Risk Assessment

The suicide risk assessment we use is the DCIB Suicide Risk Assessment. This assessment contains four elements: desire, capability, intent and buffers. This is different than our previous risk assessment, the CPR Risk Assessment.

Follow Up

An important element of working with suicidal individuals is follow up. Studies (e.g. Brown & Green, 2014) have demonstrated that follow up has the potential to reduce the likelihood of future suicide attempts and ideation.


Brown, G.K., Green, K.L. (2014) In Expert Recommendations for U.S. Research Priorities in Suicide Prevention, American Journal of Preventive Medicine. 47(3) Supplement 2:S209-S215 DOI: 10.1016/j.amepre.2014.06.006,


Cite this article as: MacDonald, D.K., (2015), "Beginnings of an Online Crisis Chat Service," retrieved on July 23, 2019 from

Facebooktwittergoogle_plusredditmailby feather

Review of Suicide Risk Assessment Applications

A short post today; while considering embarking on a side project to develop an iOS or Android application it occurred to me that suicide prevention and crisis intervention are areas where mobile applications could certainly come in handy (as well as case management.)

I downloaded an Android emulator and several free apps to prevent suicide:

  • HELP Prevent Suicide
  • R U Suicidal
  • Safety Plan
  • Suicide? HELP
  • Suicide Test

None of them were particularly useful, they all seem to follow a similar layout.

Android Suicide Prevention Applications

HELP Prevent Suicide

This application includes a link to the National Suicide Prevention Lifeline, the NSPL phone number, warning signs, some strategies for laypeople to help (asking about suicidal thoughts, listening and connecting to others) and a list of helplines. Most of the apps tended to follow this model.

R U Suicidal?

This application was unique in that it focuses on a series of videos. (There’s another version of this app for your spouse or significant other.) Recorded by a Psychologist, the intent is that you’ll watch the video, and then it pauses at points to allow you writing space for your feelings.

I found the videos un-engaging and the lack of a seek bar to move through the video or any text accompaniment means if I were in a crisis I likely wouldn’t sit through them.

Safety Plan App

This app, like the name says, just does safety plans. It gives some prompts and allows you to write information on warning signs that trigger suicidal or self-injurious thoughts, coping strategies, contacts, and so on. Because there are no defaults or examples you’d likely need a second person to help you bounce ideas off of.

Suicide Test App

The most interesting app of all was simply called Suicide Test. It had 20 yes/no questions, and then scored them to tell you your suicide risk. The font was terrible and some of the questions nearly unreadable but I liked the idea and the simple interface.

Building a New Suicide Risk Assessment Application

If I were to design a new suicide risk assessment application (and it’s on my to-do list), it would feature the following:

A comprehensive suicide risk assessment tool that meets the NSPL standards

  • A layman’s quiz to help them flesh out their own suicide risk
  • A support network/safety planning document with suggestions to help stimulate filling it out
  • Links to helplines
  • Information for both the suicidal person and the third-parties who may be viewing the app on how to reach out for help and to be non-judgemental

No application currently on the market meets these requirements, which is a real shame. Before the end of the summer I hope to have a prototype with which to begin usability testing. Watch out for it!

Cite this article as: MacDonald, D.K., (2015), "Review of Suicide Risk Assessment Applications," retrieved on July 23, 2019 from

Facebooktwittergoogle_plusredditmailby feather

Developing a Risk Management Program for Your Helpline

Importance of Helpline Risk Management

Risk management is the process of assessing and taking steps to mitigate areas where things that might go wrong in your organization. While we all hope that nothing would ever go wrong, it’s important that we are not caught blindsided. Some of these issues are within our control (e.g. helpline quality issues) while others, like natural disasters are not.

A helpful way to look at this, which came up at a conference that I attended recently, was to imagine you are being interviewed by the media because something has gone wrong. What would you tell the journalist that you currently do to avoid this problem? If the answer isn’t good enough, perhaps looking at where you can improve.

Failures in Risk Management

Failures in risk management can have wide-ranging impacts. They include a drop in morale of your volunteers, financial penalties resulting from lawsuits, a reputation hit among the community, even loss of funding and closure of your helpline.

Probability/Impact Heat Map

An interesting concept I learned in a conference was the idea of a “Probability/Impact” heat map. This was a simple way to envision the various issues that could affect your helpline. Here is an example of a simple heat map with two items, Natural Disaster and Helpline Quality Issue.


The Natural Disaster (let’s say, a hurricane) would have an extremely large impact on the helpline’s ability to operate, however hurricanes (at least here in southern Ontario) are extremely rare so their probability is low. While it might be helpful to write policies in case of a potential natural disaster (some of which are detailed in the below sections) they can likely safely be reviewed on a yearly-basis, perhaps at a Board Meeting – unless a hurricane is imminent.

On the other hand, while helpline quality issues do not have as big an impact on the helpline’s ability to operate (although still substantial, if they are widespread), they are much more common than hurricanes.

For that reason, you will likely spend more time in your day-to-day implementing policies, regular supervision and monitoring, and (ideally) having a detailed performance improvement program for volunteers identified as falling below standard. This, combined with a strong training program and an awareness in the volunteers of what is expected of them should help to mitigate issues resulting from poor quality volunteers.

As you proceed through each section, keep in mind what elements you already implement to mitigate risk and where you can improve.

Recruitment and Screening

Recruitment and screening is your first opportunity to meet potential volunteers and learn about issues that may present themselves later in their volunteerism. By taking a keen interest in the volunteers you are recruiting now, you can be better prepared for issues later.


Almost all helplines involve some kind of interview process for potential volunteers. Do you use a structured interview process, or wing it? If a potential applicant feels discriminated against during your interview process you can open yourself up to legal or public relations issues.

Risk Management for Interviewing:

  • Create a structured interview with specific questions asked of every candidate
  • When conducting interviews, write down notes for each response and store them in the volunteer’s file
  • Write a protocol explaining in what interview responses (if any) lead to supervisor review or automatic discontinuation of their application
  • Include questions on areas such as sexual assault, self-injurious behaviour, suicidal thoughts – ask volunteers if they have any exposure to these potentially “triggering” topics and discuss with applicants how they will be impacted

Police Checks

Police checks are the most obvious form of screening for helpline volunteers. In Canada, a Vulnerable Sector (VS) Check provides both a regular criminal record check, as well as information on sex crimes and other records relevant to working with vulnerable clients, even if they have received a pardon or they were not convicted.

In the UK, this is known as a DBS (Disclosure and Barring Service) check, and it works in a similar way, showing crimes that would be excluded from a normal criminal record check.

It’s important to remember that a criminal record check is not a guarantee against crime, because volunteers may become involved in the criminal justice system after the start of their volunteerism, or may have committed crimes for which they were not caught or convicted.

Risk Management for Police Checks:

  • Require police record checks for all volunteers
  • Require an updated criminal record check every several years (covered by the helpline if within your budget) or as a low-cost alternative, have volunteers sign an attestation that they have not been arrested in the previous year
  • Create a protocol for what offences you will allow on the helpline and when; for instance, you may be okay with offences that occurred more than 10 years ago, or you may only allow crimes that were not “moral turpitude” offences (drunk and disorderly may be okay, theft or assault may not be.)
  • If you allow volunteers on the lines with these crimes, document your decision to do so and have them sign a document certifying their requirement to stay “crime free” in order to continue being involved with your line

Reference Checks

Reference checks, like police checks, are common for both employees and volunteers on helplines. References allow us to learn more about a person from their peers and previous supervisors. This can be illuminating.

Reference checks should be conducted with a standardized form and specific questions looking at a potential volunteer’s interpersonal skills and how they work under stress. Examples of questions can include, “How does this person interact with other people in their life?” and “Have you ever seen this person react under stress? How did they perform?”

One misconception about references in Canada is that if an employee gives a poor reference about a colleague they are opening themselves up to legal liability. This is untrue; in fact, reference checks are considered qualified privilege (Dunlop, n.d.)

Risk Management for Reference Checks:

  • Use a standardized reference check form and record the answers; make sure to document the method of communication, the date, and the title and name of the person you spoke to
  • Ask volunteers for a minimum of two references. This shields you against problems if one reference has been “coached” to only speak positively about volunteers
  • Create a policy regarding how you will proceed if references are inaccessible; will you allow volunteers to continue during the process with only one submitted reference, will you ask them for another reference or discontinue their application?

Volunteer Training

Training is arguably the most important element to protecting callers and clients from poor quality volunteers. By providing a strong training system, volunteers are able to provide consistent service to all callers.

Emotional Reaction of Volunteers

While everyone thinks about the service to the callers, there are also risks of harming potential volunteers during training. It is important to recognize that they may be impacted or triggered by roleplays, by discussion around topics like suicide, abuse (including self-abuse) and sexual assault.

There are a number of steps you can take to protect the emotional health of your potential applicants.

Risk Management for Volunteers during Training

  • Ask volunteers who identified as undergoing counselling or therapy to provide your helpline with a letter confirming their suitability for volunteering. At Distress Centre Durham, volunteers are provided with a letter that lists the potential impacts of volunteering and clinicians are simply asked to sign the document certifying that volunteer work will pose no harm
  • Have volunteers or staff ready during the training to debrief in private with volunteers who are negatively impacted participating in roleplays
  • Closely monitoring the non-verbal signs of volunteers during discussions and documenting the results of those observations. Document everything! This way if you need to later release a volunteer from their commitment, you have evidence to support your decision

Security (Physical and Electronic)

Physical and electronic security is an important element of risk management. Many helplines are in spartan circumstances, in industrial parks or the basements of churches where there is not a lot of vehicle or person traffic. Additionally, late night shifts can have volunteers leaving at midnight, 2am, 4, 6, etc., where they could be in danger.

Risk Management for Physical Safety

  • Ensure a floodlight is available for the front door and parking lot; a peephole allowing volunteers to see out the main door will allow them to ensure the area is clear before departing to their vehicles
  • Having two volunteers on staff. By making sure that (where available) two volunteers leave and enter the building at a time, you can ensure they feel safe
  • Installing cameras or other surveillance equipment can be both a deterrent and helpful in the aftermath of a physical safety incident.
  • Multi-layered alarm systems allow you additional forms of protection. For instance, in addition to having a front door with a master key, you can install an alarm system with a code. This mean that even if your keys are compromised or the lock is picked, individuals are not able to enter your building without tripping the alarm system.
  • Establishing protocols for locking doors help volunteers ensure that they know something is amiss. For instance, if the front door is to be locked at all times, except for 5 minutes before and after a shift, a volunteer who enters the building and finds the door open at an irregular time knows to be on alert.

Informational Security

As more and more information is stored digitally, the impact of “hacking” and data theft becomes bigger and bigger. Crisis lines in particular may store a lot of information on their callers and this may be vulnerable to theft.

Risk Management for Information Security

  • Encrypt information on hard drives, using a tool like Microsoft’s free Bitlocker, so that if the hard drive or computer is stolen the information will be inaccessible without the password
  • Disconnect computers where helpline work occurs from the internet, or use a web blocker to prevent access to all non-essential websites
  • Develop information protection policies, such as locking all volunteer and caller files in keyed cabinets (and limiting access to who has the keys)
  • Consider using an online helpline management software like iCarol; all information transmitted with them is encrypted at both ends and stored offsite in secure data warehouses where security is the priority. This can help both your funders and callers feel more secure.

Natural Disasters

Natural disasters are also a concern depending on where your helpline is located. It’s important that you put in place policies to help mitigate data loss and allow your helpline to function in times of disaster, when you may find yourself a crucial source of information, emotional support and crisis intervention in the aftermath.

Risk Management for Natural Disasters

  • Develop policies on how your helpline will respond during disasters; will you forward your lines to cell phones so volunteers can take calls off-site, or will you shut down for safety?
  • Backup information on to external hard-drives or other external data sources that can be removed, or completely outsource back up to an offsite data provider so you are protected even in case of sudden disasters
  • Keep physical copies of information such as volunteer and staff contact information securely stored for emergencies

Direct Service (On the Helpline)

Sometimes risks occur from the work that your helpline workers are actually performing on the phone. These are the most worrying to volunteers, callers and managers because of their contact with often vulnerable clients.


There are a number of risk management techniques that can be put in place to help mitigate the risk poor delivery of helpline service has on your line.


Risk Management for Helpline Service Quality

  • Perform regular supervision of volunteers; this can include reading their written call reports, randomly listening to their calls live, meeting with them for weekly or monthly supervision meetings and ad hoc meetings if callers report concerns with specific volunteers
  • Establishing very clear standards and protocols around breaches of confidentiality, meeting callers offsite, giving advice, or other concerns that your helpline may face. Explain the behaviour, give examples of behaviour that falls below the standard, and explain the consequences, up to and including being released from their volunteering duties
  • Conduct regular surveys with callers. For instance, a yearly or semi-annual survey with callers can help you identify issues that may slip through the cracks if you don’t perform live call monitoring because of budgetary or ethical concerns
  • Using a tool like Chronicall can help you see patterns in call data that you may not have been aware of before. For instance, I developed a tool that allows me to input shift coverage data from iCarol and call data from Chronicall, and identify areas with high rates of shift coverage and a large numbers of missed calls. This could mean that some volunteers missed that shift, were ignoring calls or were dealing with crisis calls that required them to not answer incoming calls.


Hopefully some of these risk management techniques will have been new to your crisis line; implementation of a comprehensive risk management process in cooperation with your administration and Board of Directors is essential in mitigating potential disasters before they happen.

Do you have any recommendations for risk management methods I may not have thought of? Suggest in the comments.


Dunlop, C. (n.d.), Edmond Harden. “Qualified privilege protects comments made during reference check from defamation action”, accessed electronically from on Mar 30, 2015

Cite this article as: MacDonald, D.K., (2015), "Developing a Risk Management Program for Your Helpline," retrieved on July 23, 2019 from

Facebooktwittergoogle_plusredditmailby feather