Project Overview

Wellnicity Assessment Recommendation Engine

The founders of lifestyle startup Wellnicity, which makes top-quality supplements for a variety of common ailments such as sleeplessness, intestinal discomfort, and high cholesterol, found themselves facing a problem that has plagued the supplements industry for years: determining what supplements to recommend to patients, how much they should take, and when and how often they should take them. Crafting dosing recommendations requires clinical expertise in analyzing both lab test data and a patient’s experiential data, and combining calculations on those data with a clinician’s knowledge of the patient. This process is time-consuming and expensive to scale, a necessity for a company poised to sell to millions of customers over the Web.

So the nutritional experts at Wellnicity asked Vertex to create an expert system for supplement recommendations that would combine a patient’s lab data with and health history data from an online wellness profile to provide accurate, automated dosing recommendations. Vertex agreed to develop the system as a partner in the project and the startup eventually known as Wellnicity – and its dosing recommendation engine, now known as the Wellnicity Assessment Recommendation Engine (WARE) – was born.

How Clinicians Work with WARE

Clinicians create assessments that consist of e-health profiled and rule sets – and rules sets reference parameters, such as amounts of nutrients or other compounds, which clinicians can also define. For each assessment, the user adds a name and relevant metadata, as well as the lab associated with testing for the lab data portion of the assessment.

The clinician then creates the e-health profile, the questionnaire that provides the assessment with experiential and relevant personal data, as well as the rules for analyzing the lab results.

For the e-health profile, a question editor is provided that lets questions with a wide variety of answer types to be created (and later, evaluated according to type): free text (text line, text block); numeric (real numbers, integers, fractional integers); sets; binaries; multiple choice; date and time values; and value ranges. Almost any kind of question and answer set can be created with the flexibility of this editor.

Editing health profile questions for an assessment.

The rules that govern the evaluation of lab data for an assessment are created through a rules editor that analyzes lab results using predefined parameters (also created in the WARE system) by combining condition sets into structured rules. A single rule in the system can actually be a compound collection of “sub-rules,” and each sub rule can consist of multiple condition sets. Rules and condition sets can be nested and logical conjunctions applied; all the tools of digital logic are provided, to allow for maximum flexibility in crafting rules. Recommendations, likewise can be nested and conditional to a very wide degree.

Editing rules and creating condition sets for an assessment.

The WARE recommendation engine makes use of a simple at-home test kit and a wizard driven “e-health profile” to gather important patient data. The patient sends her test kit sample to a lab using a mailer provided with the test kit, and then registers the test kit online and takes the e-health profile assessment, which consists of a short series of questions carefully crafted in the WARE system to provide important clinical results. Once a patient’s test results are available, the lab transmits the patient data electronically to Wellnicity by means of the WARE API.

The WARE system then takes the e-health profile and lab results and uses fuzzy logic knowledge bases and the clinician-crafted rule sets to evaluate this patient data and automatically render dosing recommendations. Clinicians are notified when these assessments are complete, and can review and even modify them or leave them as is; when complete, they approve the regimen and report created for the patient and the patient is automatically notified, allowing the patient to log in to her dashboard to see her results and recommendations.

Each patient is provided a personalized report as well as a daily regimen of supplements to take and what time of day to take them. This regimen is tied to the Wellnicity Web site’s user dashboard (the site was built with Vertex CMS Site Press and Vertex also created the site’s ecommerce application – to make purchasing the recommended supplements quick and easy).

Even better, the system engages the patient in an ongoing relationship designed to promote wellness by creating a “therapeutic feedback loop,” maintaining a consistent refill program for supplements and retesting schedule for lab testing. As new customer data come in, the patient’s daily regimen is updated and modified to accommodate the changes in the customer’s clinical and experiential data. This feedback loop keeps the patient on a constant path to better wellness outcomes, and strengthens the bond the patient feels towards Wellnicity and its products.