Sometimes medical interviews based on Infermedica’s platform can last longer than expected. Why? And how can it be optimized?
Interview length will vary according to the medical issues of the end user. While seeking the most accurate list of probable conditions, our algorithm learns and adapts the interview length depending on the user’s answers.
The main aspects that determine the length of the interview are:
- Accuracy: our commitment is to deliver the most accurate result possible to every patient. We want to make sure that we consider all the relevant symptoms and deliver accurate, valid results at the end of the interview;
- Complexity of the symptoms: some symptoms might need more precision than others, and hence, more questions need to be asked about them (for example, specify which body part hurts).
Nevertheless, here are some recommendations you can follow to optimize the length of the interview:
- Encourage patients to gather their initial evidence efficiently and consider how relevant their symptoms are. A single initial symptom can determine the whole pre-medical interview. Learn more about the role of initial symptoms here ->
- Allow your patients to provide all the risk factors from the beginning of the interview. We recommend asking the patient about common risk factors before the actual interview begins. Usually you can do it on a single screen. Check out our API endpoint responsible for risk factors ->
- Use "triage" interview mode to shorten and optimize the duration of the interview. Do this if the goal of your symptom checker is to assess the correct triage level (and not to calculate the probability of the conditions). See the technical documentation for triage mode here ->
- Use Stop Recommendation. Sometimes users do not notice the “should_stop” flag and just continue with the interview until the endpoint stops returning questions. Once enough information has been collected, the interview should be stopped. To help you decide when to stop asking further questions, you should apply our stop condition recommendation. More information->
MCh, KWo