Project Background & Overview

Healthcare professionals, particularly doctors and administrative staff, face numerous challenges in managing patient care and operational workflows. Chikitsa aims to bridge these gaps by leveraging AI-powered tools to streamline administrative tasks, enhance patient outcomes, and optimize healthcare delivery. The project focused on developing intuitive, AI-enabled products, including IPD management software tailored to the needs of healthcare professionals in hospitals and clinics.

As the product designer, I spearheaded the research, user journey mapping, and design of the IPD management software and standalone AI-enabled products tailored to the diverse needs of healthcare professionals.

Research & Discovery

Research Methodology

Conducted comprehensive mixed-methods research to understand healthcare workflow challenges:

Ethnographic Studies

Spent 40+ hours shadowing healthcare professionals across 3 hospitals in Jaipur and Delhi, observing real-time workflows during patient consultations, administrative tasks, and system interactions.

User Interviews

Conducted 12 in-depth interviews (6 doctors, 3 admin managers, 2 nurses, 1 IT admin) to understand pain points, technology adoption barriers, and workflow requirements.

Diary Studies

4 doctors documented daily workflows over 2 weeks, capturing time allocation, frustration points, and system interaction patterns.

Competitive Analysis

Evaluated 3 existing healthcare platforms through product demos with their teams: Suki, Augnito, MyScribe, Simbo, Abridge, HealthPlix, Eka Care, and Driefcase. Analyzed features, workflow integration, AI capabilities, and usability patterns.

Key Research Insights

Through affinity mapping of 180+ observations, five critical themes emerged:

Time Management Crisis: Doctors spent 40% of appointment time on administrative tasks, with 15-20 minutes per patient just on documentation.

System Fragmentation: Patient data scattered across 3-5 disconnected systems requiring constant switching and manual data transfer.

Data Accessibility: Critical patient information buried in lengthy records with no quick access to medication history or diagnostic reports.

Technology Resistance: Previous implementations failed due to complexity and lack of training, creating skepticism toward new tools.

Decision Support Gap: Doctors needed data-driven insights for clinical decisions but current systems provided raw data without intelligence.

User Personas & Empathy Maps

Dr. Meera Sharma, 42

Mr. Ravi Gupta, 35

Information Architecture

Site Map

Developed hierarchical IA with 6 primary sections:

• Dashboard (personalized)

• Patient Management

• Clinical Tools

• Analytics & Insights

• Administration

• Settings

User Journey Mapping

Created detailed journey maps for primary personas highlighting pain points and opportunities.

Dr. Meera's Morning Clinic Journey:

• Phases: Pre-clinic → Consultations → Documentation → Review

• Key Pain Points: Finding records across systems, waiting for diagnostics, repetitive data entry, system timeouts

• Opportunities: Unified dashboard, predictive loading, templates, offline mode

Design Approach & Process

Framework: Double Diamond Design Process

Ideation & Prioritization

Facilitated 3 workshops generating 60+ solution concepts, prioritized using Impact vs. Effort matrix. Identified 15 high-impact, medium-effort features for MVP.

Wireframing Evolution

Low-Fidelity (Week 1-2): Paper sketches for 12 key screens
Mid-Fidelity (Week 3-4): Digital wireframes with cognitive walkthroughs
High-Fidelity (Week 5-6): Refined wireframes ready for usability testing

Usability Testing

Testing Protocol

Participants: 5 doctors, 3 admin staff
Method: Moderated sessions, think-aloud protocol
Duration: 60 minutes per session
Tasks: 8 scenarios per participant

Key Results

Task Success Rate: 87.5%
Task Completion Time: 12% faster than target
System Usability Scale (SUS): 78.5 (Grade B)
Critical Errors: 3 across all sessions

Critical Findings & Iterations

Issue 1: Analytics Discoverability
Only 37.5% found analytics without help
Solution: Moved to primary nav, added dashboard widgets
Validation: 100% discoverability in re-test

Issue 2: Prescription Efficiency
Required 4 clicks vs. expected 1-2
Solution: Combined steps, added keyboard shortcuts
Validation: Reduced to 2 clicks, 40% faster

Issue 3: AI Feature Visibility
25% missed AI insights panel
Solution: Increased prominence, added indicators
Validation: 100% recognition in follow-up

Critical Findings & Iterations

Issue 1: Analytics Discoverability
Only 37.5% found analytics without help
Solution: Moved to primary nav, added dashboard widgets
Validation: 100% discoverability in re-test

Issue 2: Prescription Efficiency
Required 4 clicks vs. expected 1-2
Solution: Combined steps, added keyboard shortcuts
Validation: Reduced to 2 clicks, 40% faster

Issue 3: AI Feature Visibility
25% missed AI insights panel
Solution: Increased prominence, added indicators
Validation: 100% recognition in follow-up

Final Design Solution

Core Features

Doctor's Dashboard

• Today's appointments with patient summaries

• Pending tasks and alerts

• AI-generated clinical insights

• Quick search and performance metrics

Integrated Patient Record

• Single-page complete history

• Timeline of interactions

• Medication reconciliation with warnings

• Allergy and contraindication highlights

AI-Assisted Prescription Tool

• Natural language input

• Automated interaction checking

• Common templates

• Digital signature integration

Administrative Command Center

• Appointment scheduling with conflict detection

• Automated billing workflows

• Patient queue management

• Compliance dashboard

Design Principles

Clinical Clarity: Information hierarchy optimized for high-pressure medical environments

Intelligent Defaults: System anticipates needs based on context, reducing cognitive load

Transparent AI: Recommendations clearly labeled with confidence levels, users maintain full control

Efficiency Through Design: Every interaction optimized for speed with keyboard shortcuts and batch operations

Results & Impact

Quantitative Outcomes

Operational Efficiency:

• 40% reduction in administrative time

• 60% faster patient record retrieval

• 75% reduction in scheduling conflicts

• 35% decrease in billing errors

User Adoption:

• 89% user satisfaction score

• 94% daily usage within 2 weeks

• 67% reduction in support tickets

Clinical Impact:

• 25% increase in patients seen per day

• 15% improvement in prescription accuracy

• 82% reported enhanced decision-making

Qualitative Feedback

Doctors:

• "This is the first EMR that actually helps me practice medicine"

• "I can finally see all patient information in one place"

Admin Staff:

• "Scheduling is so much faster and more accurate"

• "Training new staff takes days instead of weeks"

Key Learnings

Ethnographic Observation Reveals Hidden Insights: Watching doctors work revealed micro-inefficiencies they'd normalized and wouldn't mention in interviews. Real-world observation was invaluable.

Mental Models Drive Intuitive IA: Card sorting revealed how doctors and admin staff think differently about information organization. Building IA from actual mental models eliminated the need for extensive training.

AI Transparency Builds Trust: Healthcare professionals embraced AI when they understood reasoning and maintained control. Showing confidence levels and allowing overrides was critical for adoption.

Iterative Testing Prevents Costly Mistakes: Three rounds of testing and iteration caught usability issues before development, saving time and ensuring the final product met real user needs.