To calculate your matches, we analyze our secure database of 9 billion healthcare interactions:
- We search among nearly all practicing doctors in America (that’s 951 thousand doctors!)
- We factor in the things that matter to you, from the office location to the doctor’s gender.
- We rank your matches based on the number of patients like you each doctor has treated. Our algorithms assess the similarity between you and other patients by condition, gender, and age.
Our ranking algorithm is grounded in well-respected medical research and meta-reviews that have demonstrated that patient volume (how many patients a doctor treats) is typically a good predictor of medical outcomes.
For instance, Janakiraman et al. (2011) showed that the highest volume OB/GYNs had fewer complications for patients, even when adjusting for a variety of patient-specific factors. Similarly, a systematic review by Maruthappu et al. (2015) showed that more than 86% of studies found a statistically significant relationship between a surgeon’s case volume and various health outcomes such as recurrence rate, complications, and mortality.
Based on these findings, we created an algorithm that objectively ranks the doctors in your area by the number of patients like you they’ve treated. To calculate the number of patients like you a doctor has seen, our algorithm assesses the similarity between you and the doctor’s other patients using three factors:
- Reason for visit (condition or procedure): When trying to measure expertise and experience, the volume of patients a doctor has seen with your specific condition is important. Until now, people have measured expertise and experience using proxy signals: is the doctor actively researching the condition, are they affiliated with a top university, does your referring doctor think they are the best? With Amino’s comprehensive database of real patient experiences, you can now for the first time see the number of patients with your diagnosis each doctor has seen, so there’s no more second-guessing.
- Age: For some medical procedures and conditions, patient age matters a great deal in determining which doctor you’re most likely to see. For instance, children with cancer will typically see a pediatric oncologist who does not treat adults. Amino uses a mathematical technique called B-splines which are used to smooth patient ages in the cases we’ve seen. The set of B-splines we use is adaptive and changes for each condition, procedure, and specialty. For some conditions, a 32-year-old woman and a 45-year-old woman have similar concerns and will see similar doctors. But for others — like pregnancy — those women will have different concerns and may see different doctors. For each condition and specialty, Amino’s algorithms use our data to learn how similar patients of different ages are.
- Sex: Our data shows that doctors often specialize in treating either men or women — for conditions where you may expect it, like breast cancer, and also for more common conditions, like asthma. If you specify your sex when searching for doctors on Amino, we can tailor your results accordingly.
At Amino, we aim to provide you with accurate, trustworthy data. When we count the number of patients like you seen by a doctor, it’s important that we do not artificially increase our patient counts as a result of using B-splines. For example, if you specify that you’re a 32-year-old woman, we will partially count cases from younger and older women in order to create a representative picture of the number of patients like you a doctor has seen. However, we take care that each patient we see in our secure database only contributes one case in total when summed over all ages and both sexes.