Active patient engagement is critical to successful patient outcomes, but when does the interaction become too overwhelming for patients? Health systems have deployed omnichannel communications to improve patient engagement, leveraging various communication channels such as SMS text, email, and chat to engage with patients effectively. The adoption of omnichannel technology was accelerated during COVID-19 lockdowns and is here to stay. It has provided a tremendous uplift to healthcare systems by driving revenue growth and patient adherence, freeing up time for providers to respond asynchronously to patients, and improving provider safety by reducing the number of patient visits.
For all the positive impacts from a patient engagement perspective, providers must view these interactions through the lens of the post-pandemic patient. Almost overnight, well two years, everything became touchless and digitized. QR codes and apps have replaced most human interactions. The digital interaction landscape is a boon for hospitals with overstretched and exhausted providers. It provides a proven path to drive revenue through increased visits, tests, and encounters and helps keep unnecessary visits to a minimum. The thing to be mindful of is that there will be a tipping point for patients where they will begin to experience digital engagement overload. As a patient is overwhelmed with managing digital interactions, multiple touchpoints from multiple providers will become overwhelming and annoying to the patient. This could impact the effectiveness of the digital patient engagement experiment.
Just for fun, count how many patient portals you have accessed in the last year to access your patient data, set up appointments, or read provider messages. Now count how many passwords you have had to reset to access those portals. Lastly, count the number of texts you have received from various providers about appointments, survey links, upcoming vaccinations, and care reminders. Most likely, you have run out of fingers and toes at this point. Unfortunately, this is the state of the patient experience in the US these days. Each provider uses various contact points to drive patient outcomes. For someone experiencing a health issue, it is not uncommon to interact with multiple providers, inside and outside their primary health network. Each one of those providers could have a separate patient engagement framework. This creates a fragmented and frankly annoying experience for the patient.
Not only will the patient need to log into different portals to access their data, but the engagement methods will vary between providers, from offering very basic services like appointment reminders to more active engagement, such as sending message alerts, care prompts, and program information. For the patient and/or caregiver, this becomes difficult to manage. In time, the patient’s active engagement becomes less effective and turns into more digital noise that blends into the other digital noise we experience daily.
Leverage sentiment analysis framework to monitor patient experience
As the healthcare system rotates toward value-based care, measured outcomes will be the revenue drivers. The path to better outcomes will be managing adherence to the care plan. Leveraging omnichannel communications to manage patient outcomes has provided an immediate uplift for hospitals and providers. However, it is important to continually monitor patient sentiment to ensure your patient engagement framework maintains effectiveness over time.
Through formal sentiment analysis, providers can begin to capture and react to patient feedback and emotions toward provider engagement methods. The analysis should focus on capturing and measuring the patient’s impact or feeling related to the frequency, relevance, personalization, accessibility, and timeliness of provider interactions.
Frequency of interaction
Administrators defining the patient engagement framework should continue to view the interactions from the patient lens. An ill patient is dealing with the surrounding digital world and is inundated with constant reminders and information pushes, which may be a case of diminishing returns. The patient may simply become overwhelmed with the constant contact from multiple providers, and the sentiment may turn negative holistically, with the patient viewing the entire experience as negative and finding it difficult to separate the experience by a provider.
Relevance of interaction
Administrators should be careful to define patient personas based on where that patient is in the care journey and adjust the interactions accordingly. For example, a patient who has breast cancer probably does not require a reminder for a mammogram. This could come innocently from an automated reminder system but could be viewed extremely negatively by a patient who is undergoing treatment. To avoid this, it will be critical to build business rules around clinical patient data to avoid these types of unwarranted contacts.
Personalization of interaction
We are all victims of digital automation. I am sure we all have one email inbox full of emails we should have unsubscribed from years ago but failed to do so. In modern society, receiving automated, unsolicited interactions will be viewed negatively. Providers should leverage clinical data to personalize messages to the patient as much as possible. Healthcare is personal by nature. Patients are vulnerable and want to have a sense of trust in the provider. Feeling that they are receiving a sales text from Target instead of a personal message from their healthcare provider does not build trust.
Timeliness of the interaction
The timeliness of interactions is critical for any engagement framework. The interactions should be tailored to where the patient is in the care journey and focus on the next-best actions for them. This will help to extend the effectiveness of digital interaction and build trust with the patient that this is a viable method of communication. The framework can be designed to trigger alerts and send messages based on the defined care protocol, driving adherence, and providing a structure for patient and provider interactions.
Also, timeliness in responses to patient inquiries can be a sentiment measurement. Asynchronous communication is the lifeblood between the patient and the provider. The ability to leverage communication channels through text, email, or portal messages gives the patient access to the provider and flexibility for the provider to manage their schedule. The sentiment around the timeliness of provider responses can quickly turn negative. If the patient is experiencing active symptoms or feeling disconnected, this can drive a sense of distrust. It will be essential to monitor the response rate to improve patient sentiment.
Building a sentiment analysis framework
Building a robust sentiment analysis framework follows the basic framework of:
Building a base of sentiment data is a critical first step. Although surveys are the most common method to drive sentiment analysis, they can also contribute to the digital noise. Therefore, developing a robust data analytics framework may help extract sentiment based on patient feedback, such as data from social media posts, online reviews, and patient support conversations.
Leverage machine learning, text processing, and pre-trained sentiment analysis models to draw inferences and insights from the data. This will require detailed planning to understand the target analysis framed around the frequency, relevance, personalization, accessibility, and timeliness of provider interactions. A critical step in the analysis phase will be to validate the performance of the models using familiar model performance analysis methods, validating the accuracy, precision, and recall of the outputs.
The next step is to interpret the analysis to inform your actions. Continue to maintain the view of the patient and not get lost in the analysis. Does the analysis reflect a real-world experience for the patient? Are there other data points that would influence the analysis, such as the complexity of the patient’s disease, the caregiving circumstances, and Social Determinants of Health (SDOH) metrics that may be driving specific reactions or responses?
The insights gained from the analysis will help define a set of concrete actions that will help drive changes in the patient engagement model and quality improvements. The actions in this step should focus on maintaining the effectiveness of the patient engagement initiative and changing patient behavior. The goal is to foster active engagement between the provider and the patient.
Overall, providers will continue to invest in technology to drive revenue and relieve the stress on the system. Further advancements in artificial intelligence will continue to drive efficiencies in digital patient interactions. Omnichannel communication has proven to foster patient engagement and is a valuable tool for providers. However, a robust sentiment analysis framework must be implemented alongside these advancements to continually monitor the patient’s experience.
