Andean Plant Quiz: a privacy-first, evidence-based tool that recommends Andean medicinal plants and flags herb–drug interactions
Andean Plant Quiz provides privacy-first, evidence-based recommendations for Andean medicinal plants and flags herb–drug interactions with PubMed sources.
The Andean Plant Quiz brings evidence-based herb recommendations to people curious about traditional Andean remedies, pairing a short interactive questionnaire with a safety-first interaction checker so users can explore options like maca, uña de gato, and muña without handing over their data. Built as a single, dependency-free HTML file, the quiz matches a seven-question profile to a compact database of 12 species and returns ranked plant suggestions, PubMed-backed study pointers, and explicit contraindication handling — all executed client-side to preserve privacy and transparency.
Why an evidence-first plant quiz matters
Interest in botanical remedies has surged, but much of the online guidance remains commercial and opaque. The Andean Plant Quiz addresses two common user problems: signals that favor marketing over science, and the absence of clear, actionable warnings about herb–drug interactions. By tethering recommendations to published studies, linking to source identifiers, and refusing to recommend plants when severe interactions are present, the quiz positions itself as a practical triage tool rather than a promotional storefront. That distinction matters for anyone using herbs alongside pharmaceuticals, for clinicians who want a quick reference, and for technologists building trustworthy health tools.
How the quiz translates answers into recommendations
At the core of the Andean Plant Quiz is a compact matching system that converts seven user responses into a numeric "need vector" and compares it against a standardized profile for each plant. Questions cover primary health goals (for example, energy, digestion, stress relief), current medications (used only in-session to detect interactions), dietary preferences, age group, and activity level. Each plant in the database is represented by a set of weighted scores across health dimensions — an internal fingerprint that captures the strength of evidence for particular effects.
When a user completes the questionnaire, the client-side script computes similarity scores between the need vector and each plant profile. The top three plants are presented along with a confidence score that indicates how closely the match aligns with the evidence profile. Recommendations include short summaries of the primary benefits tied to specific PubMed citations and clear, prioritized warnings where relevant. The result is intended to be a starting point for further reading, not a substitute for professional medical advice.
How the interaction safety check prevents dangerous recommendations
A distinguishing feature is the interaction safety check. Before any plant reaches a results page, the quiz cross-references the user‑entered medication terms against an interaction database compiled from regulatory and peer-reviewed sources. For each plant the system checks whether documented interaction mechanisms overlap with the medication classes the user listed. If the database flags a severe interaction — for example, a plant that could interfere with immunosuppressants or act as an anticoagulant risk — the quiz excludes that plant from the recommendations entirely rather than merely burying a warning.
This approach enforces a safety-first policy: beneficial matches are shown only when they are not contraindicated for the specific medications the user reports. When milder interactions are found, the quiz surfaces them prominently with mechanism descriptions and the PubMed IDs or authoritative references so users and clinicians can assess the evidence depth themselves.
Design and technical choices: single file, zero dependencies
Rather than build the quiz with a modern framework, the project intentionally remains lightweight: a single HTML file, styled with compact CSS for animations and responsive layout, and driven by a few hundred lines of vanilla JavaScript. That minimal architecture keeps the entire dataset small — roughly 18 KB for all plant data — and enables instant loading, excellent performance, and predictable behavior across browsers and devices. The quiz also contains Schema.org FAQPage markup to improve discoverability in search without compromising the user experience.
Choosing no build step or external libraries reduces maintenance surface, removes supply-chain risk, and makes the codebase easy to audit. For short decision tools with static content and straightforward state — in this case a current question index and an answers object — adding a heavy client-side framework often increases complexity without delivering meaningful benefit.
Handling messy ethnobotanical data
One of the significant challenges when building a tool like this is the diversity and ambiguity of plant names. Common names vary by region and sometimes denote entirely different species; for example, "uña de gato" can refer to Uncaria tomentosa in one country and to other unrelated taxa elsewhere. To avoid misidentification, the quiz normalizes recommendations to binomial scientific names and links each plant profile to the original studies or authoritative taxonomies used in curation.
Curation also includes assigning confidence levels for specific indications. Early versions of the matching logic tended to promote plants that had modest support for many outcomes — producing moderate scores across the board — so the scoring was recalibrated to favor strong, specific signals over broad but weak associations. This calibration reduces false-positive recommendations and helps the tool surface plants with the clearest evidence for a stated goal.
What the quiz offers users and who should use it
The Andean Plant Quiz is aimed at curious consumers, herbal practitioners, and clinicians who want a quick, evidence-tethered starting point for exploring Andean botanicals. It is not intended as medical advice or a diagnostic instrument; instead, it functions as an educational recommendation engine that:
- Identifies plants that most closely match a user’s stated goal.
- Links each recommendation to the underlying research so readers can inspect study designs, populations, and outcomes.
- Flags and blocks plants with documented severe interactions against medications the user reports.
- Keeps the entire session local to the browser — no server-side storage, no tracking, and no email capture.
The quiz is publicly available and takes roughly 90 seconds to complete. Because it runs entirely client-side, it’s an appropriate fit for privacy-conscious users and organizations that prioritize compliance with data minimization principles.
Metrics and early user behavior
In the first weeks after launch, the quiz achieved high completion rates and meaningful interaction statistics. Most users completed all seven questions, with "energy and fatigue" emerging as the most common primary goal. Maca was the most frequently recommended plant, reflecting both its relatively broad evidence profile for energy-related claims and the algorithm’s calibration. Interaction warnings appeared in a notable share of sessions, underscoring the practical importance of checking herb–drug combinations: around one in eight users encountered at least one flagged interaction.
High completion likely stems from a combination of a clear progress indicator, concise questions, and instantaneous results. Presenting a short list of ranked options together with source citations and explicit safety information appears to sustain engagement without overwhelming users.
Developer lessons: building ethical health tools
For developers and product teams, the Andean Plant Quiz illustrates several practical lessons for building health-oriented consumer software:
- Keep the state model simple. If the entire application can be represented by a small set of variables, avoid introducing heavyweight frameworks that add cognitive overhead.
- Design for safety-first defaults. Where recommendations could cause harm, conservative behavior — such as excluding matches with severe interaction risk — is more responsible than permissive messaging.
- Make provenance visible. Providing direct links to source studies and summarizing study strength empowers users and reduces the perception of hidden claims.
- Preserve privacy by default. Running analysis client-side when feasible minimizes regulatory and ethical burdens around storing sensitive health information.
- Expect messy domain data. Invest time in normalizing names, mapping synonyms, and documenting curation decisions to reduce user confusion and auditability issues.
These principles apply beyond this quiz, to any tooling that sits at the intersection of consumer software and health information.
Broader implications for digital health and herb research
The approach taken by the Andean Plant Quiz points toward how small, focused tools can contribute to a more evidence-centered digital health ecosystem. Lightweight, auditable applications that couple curated datasets with transparent logic can improve user literacy about both the potential benefits and risks of complementary medicines. For businesses, such tools can be integrated into product pages, clinician portals, or patient education flows as a non-promotional way to surface relevant evidence.
From an industry perspective, herb‑drug interaction datasets are especially valuable. A searchable, verifiable interaction registry can be repurposed into clinical decision support, pharmacy checks, or telehealth workflows — provided appropriate validation and governance are applied. Combining that registry with privacy-first interfaces avoids some of the data governance headaches that larger systems face.
For developers building on top of such datasets, the quiz demonstrates the feasibility of delivering meaningful utility without complex back-end infrastructure. That lowers barriers for researchers and nonprofits seeking to publish evidence-based tools while maintaining accessibility.
Extending the quiz and integration opportunities
There are multiple directions for extending the concept while retaining the project’s core commitments to evidence and privacy. Potential extensions include:
- Publishing an API for the interaction database for use by clinicians or pharmacy software, subject to certification and appropriate safeguards.
- Adding optional, user-controlled export features so users can share a summary with a clinician — keeping sharing explicit and consent-based.
- Expanding the plant database and adding a versioning system that documents when profiles and interaction entries are updated and why.
- Packaging visualization components for reuse in articles or educational platforms that want to show study strengths and interaction pathways without recreating the logic.
These are natural next steps for organizations looking to convert curated ethnobotanical research into usable health tech.
Practical considerations for clinicians and product teams
Clinicians and health‑product managers considering the Andean Plant Quiz as a reference tool should note some practical boundaries. The tool is best used as a starting point in conversations, not a final source for prescribing or advising on drug regimens. It can, however, reduce the time needed to identify candidate plants and quickly surface interaction concerns that merit clinical attention. Product teams can also model patient-facing decision aids on the quiz’s principles: minimal data collection, clear provenance, and conservative safety gating.
Organizations that integrate similar tools should maintain clear documentation of curation choices and data sources, implement regular audits of interaction data, and ensure any API or integration undergoes proper security and compliance reviews.
The quiz also functions as an instructive case study for technical SEO and discoverability. Using Schema.org FAQ markup and concise on-page data keeps content indexable while avoiding heavy analytics instrumentation that would compromise privacy.
The Andean Plant Quiz is an example of how small, well-scoped interventions can raise the bar for transparency and safety in consumer-facing health software. By combining tight curation, a simple matching algorithm, and aggressive interaction filtering, it offers a practical model for turning ethnobotanical knowledge into accessible decision support without turning users into targets for marketing.
Looking ahead, the most valuable directions will be expanding the evidence base, collaborating with clinical partners to validate interaction data, and exploring opt‑in interoperability models that let users share vetted summaries with healthcare providers while keeping control of their personal information. These steps could help bridge traditional knowledge and modern clinical workflows in a way that respects both safety and cultural provenance.
