An analysis of inpatients’ and outpatients’ comments that employed artificial intelligence and natural language processing found that humor can be an effective communications device. It can help convey that the caregiver empathizes, is compassionate, is trying to be helpful, is attentive, and is pleasant. However, if humor is used in a flippant or sarcastic fashion, it can accentuate the patient’s negative perceptions about the caregiver; it adds insult to injury.
Why does humor in the workplace sometimes have magical effects and at other times is disastrous? The answer is not as simple as “some jokes are better than others,” or “some people are funny while others are not.” The reality is that injecting light-hearted remarks into professional interactions always entails risk. Guidance on when that risk is worth taking and might in fact lead to great rewards can be derived from analyses using artificial intelligence (AI) and natural language processing on data from one of the most emotionally charged professional settings: health care.
Our interest in humor in health care emerged unexpectedly during explorations of a large dataset of patient comments aimed at understanding what patients value most in their care. Artificial intelligence and natural language processing were used to extract positive and negative insights from comments in 988,161 survey responses received from patients about their experiences across the United States during 2020. Of these, 17% were inpatient and 83% outpatient experiences, respectively, and it is possible that some of the responses were from the same patients.
From these comments, our analytical technology extracted 1,270,000 insights and categorized both positive and negative insights into themes and subthemes. This technology makes it possible to identify issues that are important to patients but may not be captured by standard survey questions, such as, “How often did your doctor explain things in a way you could understand?” That question provides insight into whether clinicians conveyed information but not how they made patients feel as they did.
In fact, there are no survey questions to our knowledge in which patients are asked if they found their doctors or nurses humorous. But in our analyses of patients’ narrative comments, humor came up repeatedly when they described positive experiences with their clinicians (see the exhibit at the bottom of this article). Patients commented on how their clinicians interacted with them, and regardless of what survey questions they were asked, they seemed to see the clinician’s ability to use humor in difficult moments as reinforcing acts of caring.
Our analyses indicate that humor is not the main course when it comes to caring but is more akin to a valued condiment. The actual main course that is appreciated by patients is courtesy, respect, and related subthemes. Patients don’t remark on the technical skill of clinicians very often, but their comments suggest deep appreciation of empathy, kindness, helpfulness, and patience. And when patients note that care with these attributes was accompanied by humor, the humor seems more than welcome (see the exhibit for examples).
Using humor in this way does require intense attention to the moment and authenticity in conveying caring. This means that clinicians should focus first and foremost upon reliably being empathic and kind and trying to be helpful — i.e., by showing the behaviors characterized in the first column of the exhibit. If patients feel confident that these behaviors characterize their care, they are likely to welcome clinicians’ gentle explorations of building connection through humor.
On the other hand, when patients perceive the absence of courtesy and respect, the use of humor by caregivers adds insult to injury (see the exhibit for examples). In short, humor is not a stand-alone asset or liability. It serves to amplify the positive or the negative signals that patients pick up from their doctors and nurses.