Initial symptoms improve the quality of symptom assessments

Just one initial symptom can affect the whole medical interview. Find out why and how to work with initial symptoms

If you are planning to design or develop your own symptom-checking app or a health chatbot—stop for a moment. Imagine yourself in the shoes of a physician meeting a new patient. This patient describes a symptom he’s experiencing (such as fatigue), and you, as the doctor, need to ask him many more questions to understand his case. There are so many possible conditions that this symptom could represent, and the symptom itself is rather general.

Initial symptoms are like pieces of the puzzle of the whole health assessment picture. Illustration by Aga Więckowska.

“The majority of symptoms are nonspecific and can be present in multiple conditions” comments Otto Krawiec, GP and medical content editor at Infermedica. “With only one symptom, the physician would use a top-down approach, asking some questions that may seem irrelevant to the patient, but necessary for a physician to get on the right track.” 

Investigating each hypothesis takes a lot of time during the patient’s visit. This is valuable time that you, as the physician, also need to discuss the diagnosis and treatment. Instead, you spend it asking more and more general questions.

Meanwhile, time is running out. The patient is becoming impatient, maybe even starting to think that you’re just not competent enough.

Would this situation have unfolded differently had the patient given more initial symptoms? Probably, yes.

“Let us imagine another situation - a patient reports fatigue, weight gain, and constipation. A physician is already on the right path, as hypothyroidism has become one of the most probable conditions, while other diseases are considered as part of the differential diagnosis. Right from the start of the interview, the physician will ask more accurate questions. Our inference engine works in a similar way; by allowing and encouraging patients to input more initial symptoms, they will achieve more customized interviews”, adds Otto Krawiec.

In this regard, symptom checkers are very much like physicians. Of course, they are not going to replace doctors, but they use similar patterns to learn about the patient and guide them toward the best medical help.

What are initial symptoms in symptom checkers?

Initial symptoms refer to signs or symptoms that the user adds to a symptom-checking tool in the first steps of the symptom assessment interview. The patient then enters risk factor information (such as pregnancy, diabetes, or smoking history). The inference engine uses both the initial symptoms and risk factors to calculate the likelihood of many probable health conditions. The inference engine learns about the user through personalized questions to better understand their health and to be able to present the most probable underlying medical condition(s).

Therefore, the more symptoms a user enters into the engine at the beginning of the interview, the more customized the interview and the more accurate the final result will be.

Three initial symptoms are better than one

What about five or six initial symptoms? There is no exact number of initial symptoms that the interview should begin with. It should generally be more than one, but it ultimately depends on the given symptoms.

In the Symptomate app, users usually add 2 to 3 initial symptoms.

On the one hand, the more symptoms that are given, the more accurate the assessment of the initial symptoms will be. On the other hand, a large number of general symptoms can sometimes make the interview longer, as the engine will investigate all of them. We think the ideal number of initial symptoms is between 2 and 4. Still, the ultimate goal is to get the right balance between the number of initial symptoms and the interview length, as the latter significantly affects the user experience.

Benefits of adding more initial symptoms

The most important role of initial symptoms is to increase the quality of the interview and its results. Having more initial symptoms ensures:

  • A more personalized interview
  • Questions that are more specific to the user
  • Higher accuracy of the recommended triage
  • A better chance to get the proper medical help

For many users, these benefits translate into a more positive overall perception of the tool. A tool that provides highly personalized information to the user is more likely to be perceived as intelligent, well-designed, and reliable.

How many initial symptoms should you set?

No matter what the optimal number of initial symptoms might be, our main goal is to listen to users. It is up to them how many symptoms they experience and want to share. Some people will feel comfortable sharing all their symptoms, while others will not. In some cases, users won’t be fully sure of what they’re feeling. For those users, our /suggest endpoint may be helpful as it states other symptoms related to a given initial symptom. For everything else, questions generated by the inference engine will try to fill in the gaps.

Our role is to make the tool as accessible to the user as possible and encourage them to disclose as many initial symptoms as they want to.

Although organizing the user journey is important, in the case of initial symptoms, it is better to leave as much freedom as possible.

Example: Avoid constructing screens that ask users about only one chief complaint.

Encouraging users to add more symptoms

The examples above clearly discourage users from adding more symptoms. How then can you construct your symptom-checking tool to encourage users to share as many initial symptoms as possible? Here are some tips for each type of tool:

Initial symptoms in symptom checkers 

From the user’s perspective, symptom checkers are probably the most flexible, easy-to-use, and fast tools to investigate symptoms on the internet. By using them, users don’t have to go through countless articles and verify the information on their own. Users can choose how they want to use a digital symptom checker. On desktop or mobile—it’s up to them. Symptom checkers also give the most advanced options when it comes to asking about initial symptoms.

Start with the small things. On every screen of the symptom checker, always use the plural form of "symptoms" or "signs". This way, you are already building awareness of more than one initial symptom.

Example: In Symptomate, we refer to “symptoms” in the plural to encourage users to add more than one. We also directly ask users to “add more than one symptom.”

Avoid constructing the interview in such a way that it starts with one chief complaint. Instead, give users convenient tools to add extra symptoms. There are various ways to do this, like symptom searches, lists, or even body avatars. If possible, try to use two ways at once—users will pick their favorite. Moreover, remind your users about the importance of adding more symptoms with additional communications like: "Please try to add more than one symptom."

Check if all the ways to add symptoms work equally well on mobile and desktop devices.

Initial symptoms in medical chatbots

Although the main principles are the same as in symptom checkers, chatbots are moving us toward more conversational forms. Do your best to build a friendly and honest dialogue between the bot and the user. Start with a short greeting that presents the goals of the chat: “Hi! I’m an automatic symptom checker. I’ll try to help you find the possible causes of your symptoms.”

Then move toward the symptoms themselves. Use open-ended questions that ask your users about their symptoms: "Please describe your symptoms". "Tell me about your symptoms". "What symptoms are you experiencing today?". Use the plural form to suggest more than one symptom.

With each answer from the user, confirm if this is everything they wanted to list. A short "Do you want to add anything else?" works well.

Example: This medical chatbot asks users about other symptoms they would like to report. It also translates common terms into medical language.

Some users’ answers might not be clear for the engine as it doesn’t understand full sentences. However, our built-in natural language processing (NLP) will identify them and connect them with their respective medical terms. For example, “I have a headache” will be translated by our NLP to the “headache” symptom.

If there is no corresponding symptom in the engine, transparently communicate it with, e.g. “I’m sorry, but I don’t recognize this symptom. Can you describe it using different words?".

Initial symptoms in call centers 

Infermedica API also enables you to build tools dedicated to medical staff, like nurses or agents working at call centers. In these cases, you can use communication similar to symptom checkers or medical chatbots.

Equip your agents with troubleshooting techniques for possible difficult scenarios or answers to commonly asked questions. It’s also important to educate the team about the importance of having users provide specific and sufficient initial symptoms.

You can also add an NLP endpoint to your solutions, which recognizes conversations and marks keywords connected with symptoms or risk factors. NLP technology allows medical staff to translate common language into medical terms used in the inference engine. By doing so, a nurse or agent can review a patient’s initial symptoms to then better direct them to the appropriate care.


AKw, MMy, AHe