Enhance Comprehension of Over-the-Counter Drug Instructions for the General Public and Medical Professionals through Visualization Design

AI summary

Overview

Research area: Human-Computer Interaction / information visualization, specifically visualization design studies applied to medical communication and pharmaceutical labeling. The work sits at the intersection of visualization design study methodology, participatory design, and patient-facing health communication.

Technical level: Intermediate. The paper does not require deep technical background to follow, but it assumes familiarity with design study terminology (problem-driven iterative design, taxonomy, design requirements, pre-attentive visual encoding, and hierarchical chart types such as icicle plots, force-directed trees, cluster trees, and circle packing).

Scope in one sentence: The paper reports a 12-month visualization design study with three pharmacists that produced two tailored visual versions of an over-the-counter drug instruction (simplified for the public, complete for medical professionals), a taxonomy of OTC drug instructions sampled from the China NMPA database, a design workflow, and a controlled user study (N=60) comparing the visualized instruction against traditional text.

What This Paper Is About

Drug instructions contain information patients need to use medication safely, but they are typically dense, small-print, terminology-heavy, and hard for non-specialists (particularly older adults) to read quickly and correctly. The authors ask whether interactive visualizations can convey the same regulatory-mandated content more comprehensibly to two different audiences at once — the general public and medical professionals — who have different information needs and different levels of medical knowledge. Their goal is a reusable visualization framework, covering both the design of the instructions themselves and a way to classify which OTC drug instructions are suited to visualization at all.

Key Contributions

1. A design study of a visualization framework for OTC drugs that resulted in a prototype with two versions for different target users (a simplified version and a complete version), produced through an iterative process with three pharmacists over a 12-month period.

2. A taxonomy of drug instructions developed from a systematic classification of OTC drugs sampled from the official NMPA database, which received positive expert feedback. The taxonomy identifies which instruction types are amenable to visualization and which are not.

3. A design workflow for visualizing OTC drug instructions, distilled from the design exploration and user study feedback, intended to be generalizable to other OTC drug instructions.

4. A controlled user study (N=60) evaluating a visualization version of a frequently used OTC drug instruction against the traditional text version, measuring response time and usability.

The authors state that, to the best of their knowledge, this is the first attempt at systematic visualization design for drug instructions as a whole while considering different populations, and that their design, taxonomy, and workflow are generic enough to be potentially adapted to drug instructions worldwide.

Main Findings

• Visualized instructions outperformed text on speed and usability: The controlled user study with participants from the general public (N=60) found that the prototype had a shorter response time and a higher usability rating compared to traditional text-based instructions.

• Offering two versions was itself beneficial: Beyond the comparison against text, the availability of two versions (simplified for the public, complete for professionals) was found to be beneficial.

• Prescription drugs are a poor fit for this kind of visualization: An initial attempt to visualize levofloxacin tablets (a prescription drug) failed because its indications involve diseases that non-experts struggle to understand, and its usage and dosage vary by indication. The team concluded the value of visualized instructions for prescription drugs may be limited to medical professionals and is not an urgent requirement for all stakeholders.

• A three-element tuple underlies the taxonomy: Instructions are characterized by C = (u, i, p) where u is usage and dosage (changes, fixed), i is indication (diseases, symptoms), and p is population or age (specific, general). This yields 8 possible combinations, which the authors reduced to three unique cases after grouping disease-based indications together and noting that (u: fixed, i: symptoms, p: specific) is equivalent to (u: fixed, i: symptoms, p: general).

• Five elemental categories of drug instructions: Cat. 1 — usage and dosage vary with indications (mainly symptoms); Cat. 2 — usage and dosage vary with age/population; Cat. 3 — usage and dosage are fixed; Cat. 4 — indications include diseases; Cat. 5 — prescription drugs. The paper focuses its visualization work on Cat. 1–3, the OTC instructions whose indications are symptoms. Some drugs fall into more than one category (e.g., Cat. 1 + 2).

• Database scale and sampling: The NMPA database examined contains 733 Class A and 392 Class B drugs (classified by safety level). Because instructions had to be read manually, the authors sampled at a 10% rate: 73 Class A and 39 Class B drugs, totaling 112 OTC drugs.

• Section priority measured quantitatively: In a workshop prioritization activity, "Indications" and "Dosage and Usage" received the highest average scores (>4.0), followed by "Precautions," with pharmacists uniquely emphasizing "Drug Interactions."

• Visual encoding choices that made it into the final design: Dosage frequency ranges (e.g., 3–5 times per day) are shown with solid plus dashed rectangles (solid = confirmed, dashed = optional/possible); precautions are recoded into three severity categories — contraindication, not recommended/not allowed, and use with caution — plus a "discontinue the drug and consult a physician" case; drug interactions use an icicle plot after four alternatives (icicle plot, force-directed tree, cluster tree, circle packing) were sketched and discussed; pre-attentive highlighting (red font and dotted orange boxes) marks content that changes when the user switches indications.

• Legends were redrawn after expert critique: Following feedback, the precaution legends were revised, inspired by USP pharmaceutical pictograms and traffic signs, to a circle with a slash (contraindication), a triangle (use with caution), and a rectangle with a slash (not recommended/allowed).

• Pharmacist feedback changed the interface in concrete ways: Suggestions included expanding the information inside rectangles to clarify method, frequency, and duration; using pre-attentive elements to flag changes; using an icicle-based visualization for the "discontinue and consult a physician" scenario for consistency with drug interactions; placing professional information below core content; avoiding drug-specific packaging imagery for copyright reasons; renaming navigation labels from "patient" and "pharmacists" to "simplified" and "complete"; removing the double-ring ingredient chart as space-consuming without much comprehension benefit; removing the medication-methods module because it does not change by drug; and resolving contradictions such as "once every 8 hours" versus "three times a day." Pharmacists emphasized that all information must remain legally accurate without oversimplification.

• The framework was adapted to two further categories: With pharmacist input, the authors applied the framework to cetirizine dihydrochloride tablets (Cat. 2, dosage varies with age/population) and ambroxol hydrochloride dispersible tablets (Cat. 3, fixed dosing). The core adaptation across categories is how the usage and dosage section visually links to indications or populations; precautions and drug interactions follow consistent patterns with only content varying.

Methodology in Plain English

The team ran what visualization researchers call a design study: rather than building a tool in isolation, they partnered with domain experts to solve a real-world problem. Three pharmacists with 12, 13, and 13 years of pharmaceutical experience collaborated throughout, and the authors' interdisciplinary team contributed clinical medicine, pharmacoepidemiology, health data science, visualization, communication, HCI, and accessibility expertise.

The process unfolded in stages:

1. A failed first attempt that set the direction. The pharmacists picked levofloxacin tablets as a typical clinical drug. The team tried to visualize its instruction and found they could not convey its indications, usage, and dosage meaningfully to non-experts. That failure directly caused the decision to focus on OTC drugs instead of prescription drugs.

2. Building a taxonomy. Because it was unclear which instructions could be visualized at all, the team defined the (usage/dosage, indication, population) characterization described above, sampled 10% of the NMPA OTC database, manually classified the 112 sampled drugs with two researchers independently cross-validating the classification and resolving discrepancies by discussion, and condensed the result into five categories.

3. Eliciting design requirements. A structured participatory design workshop with the three pharmacists used paracetamol sustained-release tablets as a case study, guided by a tailored medical visualization model and facilitated with Five Design Sheets (FDS) for ideation and VizItCards for rapid sketching and critique. Activities included brainstorming, low-fidelity sketching, card sorting, and Likert-scale scoring of section importance. These outputs were consolidated into four requirements: R1 (two audience-specific views), R2 (layout by section priority), R3 (explicitly correlate usage/dosage with indications and populations), and R4 (match visualization technique to content type).

4. Iterative prototyping and critique. The team built an initial interactive prototype, gathered a comprehensive feedback session from the same three pharmacists covering visual encoding, layout, and content accuracy, and revised the prototype accordingly.

5. Evaluation. The final prototype was tested with potential target users from the general public in a controlled user study (N=60), comparing it against traditional text-based instructions on response time and usability. The paper content provided here describes the study's outcome at the level reported in the abstract but does not include the detailed protocol or statistics.

Methodologically, the authors state they were primarily influenced by the design study methodology and also incorporated elements of classical user-centered design and participatory design.

Why This Matters

Impact on research. The paper argues that prior work on pharmaceutical comprehensibility focused on isolated pieces — individual pictograms, dosage charts, timing visualizations — rather than treating the drug instruction document as a whole. It positions drug instruction communication as a "wicked problem" comparable to patient data communication, where different stakeholders (patients, doctors, pharmacists) disagree about needs and priorities. The taxonomy and workflow give other researchers a structured way to decide what to visualize and how, and the work extends the visualization-for-medical-communication literature from disease states and health trends into regulatory-mandated, standardized documents.

Real-world applications:

• Consumer-facing medication labels and leaflets, especially for older adults, who the paper notes face particular difficulty understanding drug instructions.
• Digital pharmacy and patient-portal interfaces that need to show dosing that changes by symptom or by age group, while keeping mandated text available on demand.
• Regulatory and standards work, connecting to existing initiatives the paper cites, such as guidelines for simplified and large-print drug instruction versions plus an electronic "complete version," and the U.S. Patient Medication Information (PMI) Proposed Rule aimed at simplified, standardized, accessible formats.
• Pharmacist counseling and doctor-patient communication, particularly for dosage, precautions, and drug interactions, which pharmacists identified as a high-risk area for misunderstanding in text.

Industry relevance. Any organization that produces or distributes OTC medication information — pharmaceutical manufacturers, pharmacy chains, health app developers, and regulators such as the U.S. FDA, the European Medicines Agency (EMA), and China's NMPA — faces the same constraint the paper highlights: information must remain legally accurate and not oversimplified, yet be comprehensible to non-specialists. The paper's two-version design directly addresses the tension between those two demands.

Future Directions

• Extending the framework to Categories 4 and 5. The authors deliberately excluded instructions whose indications are diseases (Cat. 4) and prescription drugs (Cat. 5), concluding that disease-based indications are difficult to convey with visualization and that visualized prescription instructions may only be valuable to professionals. Whether a different visualization strategy could serve those categories remains open.

• Evaluating the framework across all three OTC categories with users. The controlled user study is described as evaluating "a specific OTC drug instruction," while the framework was also adapted to cetirizine (Cat. 2) and ambroxol (Cat. 3). Whether the adaptations for those categories perform as well in user testing is not reported in the content provided.

• Testing with medical professionals as participants. The reported user study drew participants from the general public (N=60), even though the design includes a complete version intended for medical professionals. The professional-facing version's effectiveness is not reported.

• Localization and cross-national generalization. The paper shows paracetamol/acetaminophen instructions differing across China, the US, and the UK, and claims the design, taxonomy, and workflow are generic and potentially adaptable worldwide. Validating that claim against other regulatory formats and languages is a natural next step.

• Moving from prototype to deployable tool. The authors note the interactive visualization tool link will be provided upon acceptance, and that the prototype was refined through two all-hands meetings with pharmacists plus numerous minor feedback rounds. Longer-term field use and maintenance of legal accuracy across drug updates is an open practical question.

Target Audience

This paper is most useful to visualization and HCI researchers working on design studies and health communication, particularly those interested in stakeholder-driven design for heterogeneous audiences. It is also valuable to health communication and pharmacy informatics researchers, regulatory and drug-labeling professionals, and designers of patient-facing digital health products who need to present mandated medication information in accessible forms. Pharmacists and clinicians interested in how instruction design affects comprehension will find the requirements elicitation (R1–R4) and the pharmacist critique of the prototype directly relevant. The paper is written at an intermediate level: readers without a visualization background can follow the problem, the requirements, and the evaluation outcome, while the detailed chart-type comparisons (icicle plot, force-directed tree, cluster tree, circle packing) will be most meaningful to those with some visualization design familiarity.

Authors’ abstract

Drug instructions are crucial for guiding the rational use of medication. We conduct a visualization design study to enhance the comprehension of over-the-counter (OTC) drug instructions, targeting both the general public and medical professionals. We devise two tailored drug instruction designs for different audience groups through an iterative design process. A controlled user study reveals that our design outperforms traditional text-based instructions in terms of response time and usability, and the availability of two versions is also found to be beneficial. This study also motivates a taxonomy based on a systematic classification of OTC drug instructions sampled from an official drug database, which received positive expert feedback. Finally, this study summarizes a workflow for a visualization design strategy based on our design exploration and user study feedback, which can be generalized to other OTC drug instructions.