I'm a big fan of interacting with help desks or service desks or whatever you want to call them. I love having problems during my workday which prevent me from getting my stuff done, reaching out to a group of complete strangers, and having them ask me questions that may or may not be relevant to the issue at hand. Sometimes these friendly folks can help me fix my concern (in the industry, we call this first call resolution because we are fancy and sophisticated!). Other times, I will need to engage with a particular expert to help me get the answers I need. Either way, there is nothing I would rather do than spend time with a help desk. Yet even I, the self-styled Great Complainer, acknowledge how important this offering is, how lost I’d be without it, and how much more it can do to move the needle on clinician burnout.
If you sense some sarcasm at the beginning of this post, you must be a frequent reader of my blog! Am I really a help desk hater? Absolutely not! Working at a service desk is often a difficult and thankless task. Most folks seeking help just want to get back to their day, so they are not as friendly or understanding as they might be otherwise. Further, callers to a help desk can range from the “How do I reboot my computer?” folks to “I just need you to reconfigure ETX 3097767 so that the restrictor preventing it from showing up for Cardiology is removed, ok?” sorts of people. One never knows what awaits a support desk professional on the other end of the phone or text bubble.
We will always need humans to be involved in some part of the support process. However, there is of course a role for technology. For straightforward problems that can often be resolved in a few steps, chatbots serve as an excellent first step (“pre-help desk”). If the conversational bot can point a user to a training document or even answer the question at hand, everyone is happy. Often, the bot can get enough information to properly classify the problem so that if a human is needed for resolution, the proper human can be identified right away to avoid multiple handoffs and referrals. Bots that are based on artificial intelligence can get “smarter” as they learn about an organization, its particular workflow patterns, and how its staff interacts with the service desk.
One technology feature that I believe is woefully underused in healthcare is the help desk button built into many electronic health records (EHRs). I guess I should not say “built into EHRs” but rather “can be built into EHRs.” Many vendors have functionality that allows a user to click on a button in the EHR itself if they are having problems. The best of these tools does more than make it easy to send a message to a bot or a human without having to leave the EHR. Ideally, when the button is clicked, the software collects all kinds of behind-the-scenes information that can be incredibly useful to analysts when trying to solve problems. One vendor I have worked with even takes a screengrab when the button is pressed; you can imagine how helpful a picture of the screen is to a problem solver. It is said that a screengrab is worth a thousand words.
So, if this help desk button is so great, why are so few healthcare systems making it available as part of their very expensive electronic health records? In my experience, the major reason is that it is often not easily integrated into third-party service desk systems. Without a seamless connection between the EHR and the help desk software, support folks and analysts might have at least two different places to receive and deal with tickets. While that might just seem annoying, it is really a showstopper. We need a single source of truth for staff’s questions and issues, and having multiple systems to deal with makes it much more likely that follow-up will be significantly more difficult.
Once the integration is set up between the EHR and the support ticketing software, we are off to the proverbial races. If we make it easy for our colleagues to give us feedback and identify problems, they will do just that. However, as Shakespeare wrote, “Ay, there’s the rub.” If we make it easy to send us problems, we should expect more problems to come to our attention. Many of us see what we consider to be minor sub-optimal processes or setups all day long, but we rarely bring them to anyone’s attention because it seems petty or unimportant. Yet these minor issues add up to bigger problems when they affect many people. For instance, I was working with a hospital system and observed a nurse rooming a patient in the child neurology clinic. When asked about the reason for the visit, the mom noted that her daughter was having headaches. The nurse then went to a discrete list in the EHR of common reasons for the visit and selected “Other.” This astounded me. I asked her why she didn’t choose “headache.” She told me that “headache” was not an option. Hello! That is a clear oversight in the build for a child neurology clinic. I asked the nurse why she never reported that to the build team so they could add it. Her reply: How would I get that information to them? How indeed!
While I do not think it is reasonable to expect fixes to every identifiable problem, I believe that just acknowledging the problem (and the person who is calling it out) is a big step in the right direction. Some problems are simple (e.g., there is a typo in this documentation template); some problems are virtually unsolvable (e.g., the system should just know what I want to do all the time). However, even without fixing problems, most folks are just happy to be acknowledged and given an honest answer as to chances for resolution. (For an excellent explainer of the levels of EHR build options, see UC Health CMIO Dr. CT Lin’s one-pager.)
Once you have made it easy to do the right thing (i.e., submit a support request), encourage your clinicians to send problems other than just those that are technical. For instance, many specialists have good reasons for wanting specific sorts of patients to be scheduled in certain clinics on particular days because they will need equipment or staff not available everywhere. When a scheduling snafu occurs, there is typically no easy way to get feedback to the responsible staff or supervisor for future instances. This is a great use for a help desk button!
The help desk should be considered a strategic asset that can be used not only to quickly resolve the problems of busy clinicians but also to make technology, operations, and facilities function better by fixing small, annoying problems before they morph into major considerations.