Millions of patients come to Zocdoc each month to find and book appointments with healthcare providers across more than 50 specialties. Throughout the past ten years, our users have booked appointments for more than 1,000 different procedures – from annual physicals and physical therapy to cardiology screenings and counseling.
This is no small feat, considering the healthcare industry uses academic terminology and alphabetized lists of hard-to-spell specialties to categorize and deliver care. At the same time, according to the CDC, nine in ten Americans struggle to understand jargon-filled health information. This puts patients at a disadvantage, making it difficult to decipher what type of appointment they need or which specialist they should see to treat a specific symptom or condition.
This disconnect between medical speak and patients’ own colloquial language – think “gyno” not “obstetrician-gynecologist” – can frustrate, intimidate or even deter patients in their searches. This is especially concerning at the critical moment when patients are motivated to seek care.
Introducing Patient-Powered Search™
Zocdoc’s new Patient-Powered Search™ is a more intuitive search experience, built specifically to bridge the gap between healthcare industry and human speak. With Patient-Powered Search, each person can use his or her own language – from “gyno” to “hurt wrist” to “post-election stress disorder”– to confidently find the right provider for their needs. We’ve even peppered in some fun emoji searches, like ?? for allergies, ❤️? for heartburn, and ✈️? for travel medicine.
This is a far cry from traditional physician search systems, which put the onus on patients to know the appropriate specialty for a procedure or the precise medical terminology to describe their condition. In fact, that’s what we started with a decade ago. Our previous search experience followed industry convention, which meant patients had to select a specialist from a lengthy drop-down menu, then select a procedure or visit reason from another pre-selected list. In other words, they needed to know exactly what they were looking for, in healthcare’s terms. This was working for millions of patients, but we knew we needed to evolve our search experience to one that was more dynamic, intuitive, and oriented around the patient.
After a beta period on desktop, we’re now rolling out the new experience to mobile web and mobile app this week. As we like to say, with Patient-Powered Search, you no longer have to be a doctor to find a doctor.
Understanding Patient Intent Through Smart Technology
In healthcare – more so than in travel, retail or other online research – mapping what the user enters to the right results is critically important. Likewise, a smart search experience also needs to be able to learn and evolve over time from the inputs users enter. While Patient-Powered Search doesn’t look that different from search experiences found across the web, we’ve combined smart technology that understands human language with insights from a decade of past data on appointment bookings to build a product that puts more control in patients’ hands.
In building Patient-Powered Search, our goal was to use data-driven technology to power a search engine that will evolve over time. The team started by building a machine learning algorithm using existing medical literature gathered online. This algorithm is a natural language processing model, which means it processes and understands the ways humans communicate, and then maps those colloquial terms to the appropriate specialty, visit reason or procedure type. We call this the semantic search technology.
We then coupled this semantic search with string-matching technology to surface the most relevant results, and to account for common variations like potential misspellings or slang. For instance, we’re able to map the misspelled “hemroids” to hemorrhoids and “gyno” to OB-GYN based on both the strings of letters associated with each word and our understanding of the patient’s intent. Additionally, during our beta period, some of the most common searches were simply put symptoms, conditions or even body parts, such as “stomach,” “TMJ” and “tonsil,” which, thanks to our algorithm, now successfully map to gastroenterologists, prosthodontists and ENTs.
We are also using machine learning to aggregate intent, and prioritize and re-rank search results based on patient behaviors so the most popular search selections will appear at the top. This allows us to continually learn what patients are looking for, and adjust results based on trends or new needs. In other words, patient searches are truly powering the feature – the more searches and selections we collect, the better our results will get.
Giving Power to the Patient
We leveraged the insights we’ve gathered from millions of users over ten years to build Patient-Powered Search, and new searches will help us continue to improve. On top of that, we regularly work with medical experts and our data science team to review and ensure our mapping is as accurate and comprehensive as possible. And because we can understand which search behaviors ultimately lead to completed appointments and positive reviews, we will continually learn and evolve based on the entire patient journey.
Patient-Powered Search lays an important foundation for the future – it gives Zocdoc, our patients, and our partners an unmatched advantage when it comes to healthcare search, discovery and booking. And, most importantly, it furthers our core mission to give power to the patient – putting patients’ needs at the center of our product innovation to remove friction throughout their healthcare journey.
This is just the beginning: Patient-Powered Search will get smarter as more searches are entered. Help millions of patients find the care they need by adding to new terms to our database – click through to add how you’d describe a few common conditions, like anxiety, acid reflux, and more to the sample ideas provided – and contribute to our learning curve!