Research Synthesis Guidelines

Research Synthesis Guidelines

Systematic framework for documenting, synthesizing, and grading research evidence for MYCURE healthcare products with emphasis on Philippine healthcare context.

When This Skill Activates

• Documenting user research findings
• Synthesizing market analysis or competitive research
• Grading evidence quality and confidence levels
• Writing research reports or insights documents
• Analyzing Philippine healthcare systems (LGU, FHISIS, PhilHealth)
• Conducting stakeholder interviews or field studies
• Evaluating healthcare workflows and pain points

Core Principles

1. Evidence-Based Decision Making

All product decisions must be grounded in research.

Why:

• Healthcare products impact patient care
• Assumptions can lead to dangerous design flaws
• Philippine healthcare context has unique requirements
• Stakeholder buy-in requires credible evidence

2. Transparent Confidence Grading

Every finding must be labeled with confidence level.

Why:

• Not all evidence is equal quality
• Teams need to know which findings to trust
• HIGH confidence findings drive major decisions
• LOW confidence findings require further investigation

3. Triangulation for Validation

Multiple sources strengthen findings.

Why:

• Single sources can be biased or incomplete
• Triangulation reduces error
• Philippine healthcare varies by region (urban vs. rural)
• Cross-validation ensures accuracy

Confidence Grading Framework

HIGH Confidence

Definition: Finding is well-supported by multiple high-quality sources with consistent evidence.

Requirements (need ALL of these):

• ✅ 3+ independent sources confirming the finding
• ✅ Primary research included (interviews, observations, surveys)
• ✅ Recent data (within 2 years for healthcare)
• ✅ Philippine-specific evidence (not extrapolated from other countries)
• ✅ Consistent across all sources (no major contradictions)

Example:

## Finding: Manual registration takes 12-15 minutes per patient [HIGH CONFIDENCE]

**Evidence:**
1. **Direct observation** - Observed 15 patient registrations at 3 Manila clinics,
   average time 14.2 minutes (June 2024)
2. **Stakeholder interviews** - 8 clinic administrators reported 12-15 minute average
   registration time (May-June 2024)
3. **DOH data** - Department of Health study cited 13.5 minute average for
   Metro Manila clinics (2023)

**Triangulation:** Primary observation + stakeholder reports + government data = HIGH

MEDIUM Confidence

Definition: Finding has some support but gaps in evidence quality, recency, or triangulation.

Characteristics (one or more):

• ⚠️ 1-2 sources (not fully triangulated)
• ⚠️ Secondary research only (no primary data collected)
• ⚠️ Dated evidence (2-5 years old)
• ⚠️ Extrapolated from similar contexts (other SE Asian countries)
• ⚠️ Minor inconsistencies between sources

Example:

## Finding: LGU health workers prefer mobile apps over desktop [MEDIUM CONFIDENCE]

**Evidence:**
1. **Industry report** - 2022 Southeast Asia digital health survey showed 68%
   preference for mobile in rural health settings (regional, not PH-specific)
2. **Anecdotal** - 2 RHU staff mentioned preferring mobile during informal
   conversations (small sample)

**Gaps:** No Philippine-specific data, small sample size, older regional data

**Recommendation:** Conduct targeted survey of Philippine LGU health workers before
making mobile-first decision

LOW Confidence

Definition: Finding is speculative, unsupported, or based on weak evidence. Requires further investigation.

Characteristics (one or more):

• ❌ Single source or anecdotal only
• ❌ No primary research
• ❌ Outdated (5+ years old)
• ❌ Not Philippine-specific (US/EU data extrapolated)
• ❌ Contradictory evidence exists
• ❌ Assumption not validated

Example:

## Finding: Clinics willing to pay ₱50,000/month for MYCURE [LOW CONFIDENCE]

**Evidence:**
1. **Assumption** - Based on perceived value, not actual willingness-to-pay data
2. **No validation** - Have not asked clinics about pricing

**Status:** HYPOTHESIS ONLY - Requires pricing research before proceeding

**Next steps:** Conduct pricing sensitivity survey with 20+ target clinics

Triangulation Methodology

Triangulation = Using multiple data sources or methods to validate findings.

Types of Triangulation

1. Data Triangulation (Most Common)

Combine different data sources:

Sources:

• Primary research: Interviews, observations, surveys you conduct
• Secondary research: Published studies, reports, government data
• Internal data: Usage analytics, support tickets, sales feedback

Example:

Finding: Inventory mismanagement costs clinics 15-20% of medication budget

**Data triangulation:**
1. **Primary:** Interviewed 10 clinic administrators, 8 reported 15-25% loss
2. **Secondary:** DOH 2023 report cited 18% average medication wastage
3. **Internal:** MYCURE pilot clinic reduced waste from 22% to 5% with inventory tracking

**Result:** HIGH confidence - three independent data types align

2. Method Triangulation

Use different research methods:

Methods:

• Interviews (qualitative depth)
• Surveys (quantitative breadth)
• Observation (behavioral truth)
• Analytics (usage patterns)

Example:

Finding: Receptionists skip validation fields to save time

**Method triangulation:**
1. **Interview:** Receptionists admitted skipping non-required fields "to move faster"
2. **Observation:** Watched 5 receptionists skip 70% of optional fields
3. **Analytics:** Form completion data shows optional fields filled <30% of the time

**Result:** HIGH confidence - stated behavior matches observed and measured behavior

3. Philippine Healthcare Context Triangulation

Validate across different Philippine healthcare settings:

Settings:

• Private clinics (urban, well-resourced)
• LGU health centers (rural, limited resources)
• Government hospitals (public sector)
• RHUs/BHSs (barangay-level primary care)

Example:

Finding: Internet connectivity is unreliable for healthcare IT systems

**Context triangulation:**
1. **Urban private clinics:** Reliable fiber connection, 99% uptime
2. **RHUs (rural):** 2G/3G only, frequent outages, <50% reliability
3. **LGU city health offices:** DSL or fiber, 80-90% reliability

**Result:** MEDIUM-HIGH confidence - varies by setting, MYCURE must work offline

Citation Standards

Why Citations Matter

Credibility:

• Allows verification of claims
• Shows rigor and thoroughness
• Enables future researchers to build on work
• Required for stakeholder trust

Citation Format

Use this structure:

## Finding Title [CONFIDENCE LEVEL]

**Summary:** [1-2 sentence finding statement]

**Evidence:**

1. **[Source Type]** - [Name/Organization], "[Title or Description]",
   [Date], [Specific data point or quote]

2. **[Source Type]** - [Name/Organization], "[Title or Description]",
   [Date], [Specific data point or quote]

3. **[Source Type]** - [Name/Organization], "[Title or Description]",
   [Date], [Specific data point or quote]

**Triangulation assessment:** [How sources validate each other]

**Limitations:** [Any gaps, biases, or caveats]

**Implications:** [What this means for product/design decisions]

Source Types

Interviews:

**Interview** - Dr. Maria Santos (General Practitioner, Manila Clinic),
"Patient registration currently takes 15 minutes on average",
June 15, 2024, Stakeholder interview

Observations:

**Direct observation** - Field study at 3 Metro Manila clinics,
14.2 minute average registration time (n=15 patients),
June 10-12, 2024, User research

Documents:

**Government report** - Department of Health Philippines,
"Philippine Health Facility Survey 2023", Published March 2023,
Page 42: "Average patient wait time 45 minutes in public health centers"

Surveys:

**Survey** - MYCURE target market survey (n=50 clinic administrators),
"73% report using paper-based records as primary system",
Conducted May 2024

Analytics:

**Usage data** - MYCURE pilot program analytics (n=5 clinics, 6 months),
"Registration time reduced from 15min to 5min average",
Jan-June 2024

Websites:

**Online source** - PhilHealth official website,
"PhilHealth Claims Processing Guidelines",
Accessed June 20, 2024,
URL: https://www.philhealth.gov.ph/claims/

Research Report Template

---
title: "[Research Topic]"
research_type: User Research | Market Analysis | Competitive Analysis | Field Study
date_conducted: 2024-06-15
researchers: [Names]
status: draft | final
confidentiality: internal
---

# [Research Topic]

## Executive Summary

[3-5 bullet points of key findings with confidence levels]

- Finding 1 [HIGH CONFIDENCE]
- Finding 2 [MEDIUM CONFIDENCE]
- Finding 3 [HIGH CONFIDENCE]

---

## Research Methodology

**Type:** [User interviews | Surveys | Field observations | Document analysis]

**Sample:**
- Size: [Number of participants/sources]
- Description: [Who/what was studied]
- Selection: [How sample was chosen]

**Dates:** [When research conducted]

**Location:** [Where research took place]

**Limitations:**
- [Limitation 1]
- [Limitation 2]

---

## Findings

### Finding 1: [Title] [HIGH CONFIDENCE]

**Summary:** [1-2 sentence description]

**Evidence:**

1. **[Source type]** - [Citation]
2. **[Source type]** - [Citation]
3. **[Source type]** - [Citation]

**Triangulation:** [How sources validate finding]

**Implications:** [What this means for MYCURE]

---

### Finding 2: [Title] [MEDIUM CONFIDENCE]

[Same structure...]

---

## Recommendations

### Immediate Actions (HIGH Confidence Findings)

1. **Recommendation 1** - Based on Finding 1
   - Action: [Specific next step]
   - Owner: [Who should do it]
   - Timeline: [When]

### Further Investigation Needed (LOW/MEDIUM Confidence)

1. **Research Gap 1** - Finding 3 needs validation
   - Method: [How to validate]
   - Timeline: [When to conduct]

---

## Appendices

### Appendix A: Interview Guide
[Interview questions used]

### Appendix B: Raw Data
[Survey results, observation notes, etc.]

### Appendix C: Sources
[Full bibliography of secondary sources]

Philippine Healthcare Context

Key Systems to Understand

FHISIS (Field Health Service Information System):

• DOH reporting system for LGU health centers
• Monthly reporting requirements
• Specific data formats and fields
• Research implications: MYCURE must support FHISIS export

PhilHealth:

• National health insurance
• Claims processing requirements
• Accreditation standards
• Research implications: Integration needs for billing

LGU Health Structure:

• Provincial Health Office (PHO)
• City/Municipal Health Office (CHO/MHO)
• Rural Health Units (RHUs)
• Barangay Health Stations (BHS)
• Research implications: Varying resource levels, connectivity, literacy

Research Considerations

Urban vs. Rural:

• Infrastructure varies dramatically
• Internet connectivity: Urban 90%+ vs. Rural 30-50%
• Staff technical literacy: Urban high vs. Rural mixed
• Don't extrapolate urban findings to rural contexts

Public vs. Private:

• Resource levels differ significantly
• Public: Government budget constraints, high volume
• Private: Better resources, lower volume
• Research both sectors separately

Regional Variations:

• Metro Manila ≠ Provinces ≠ BARMM
• Language: English/Filipino in NCR, regional languages in provinces
• Validate findings across multiple regions

Quality Checklist

Before finalizing research:

• All findings graded (HIGH/MEDIUM/LOW)
• HIGH confidence = 3+ sources triangulated
• Citations complete with dates and specifics
• Philippine-specific data (not extrapolated)
• Recent evidence (within 2 years preferred)
• Methodology documented (reproducible)
• Limitations acknowledged (no overclaiming)
• Recommendations actionable (tied to findings)
• Executive summary for stakeholders
• Raw data preserved in appendices

Common Pitfalls

❌ Don't Do This

Treating all evidence equally:

Finding: Clinics want mobile-first design

Evidence: Read blog post about mobile trends

Confidence: HIGH ← WRONG

Extrapolating without validation:

Finding: Philippine clinics will behave like US clinics

Evidence: US healthcare IT study

Confidence: MEDIUM ← WRONG (should be LOW)

Ignoring contradictory evidence:

Finding: All clinics prefer cloud-based

Evidence: 5 clinics said yes
[Ignored: 3 clinics said no due to connectivity]

Confidence: HIGH ← WRONG (cherry-picking)

✅ Do This

Grade honestly:

Finding: Clinics may prefer mobile-first design

Evidence: General mobile trends blog post (not healthcare-specific)

Confidence: LOW

Next step: Conduct targeted survey of Philippine clinics

Validate locally:

Finding: Philippine LGU health workers face connectivity challenges

Evidence:
1. Interview: 10 RHU staff in Bulacan reported frequent outages
2. Observation: Visited 5 RHUs, saw 3G-only connectivity
3. DOH report: 2023 study cited 40% of rural health centers lack reliable internet

Confidence: HIGH (triangulated, Philippine-specific, recent)

Summary

Research synthesis framework:

1. Grade every finding: HIGH/MEDIUM/LOW confidence
2. Triangulate: 3+ sources for HIGH confidence
3. Cite rigorously: Enable verification
4. Context matters: Philippine healthcare is unique
5. Be honest: Acknowledge limitations and gaps

Confidence requirements:

• HIGH: 3+ sources, primary research, recent, PH-specific, consistent
• MEDIUM: 1-2 sources, some gaps in quality/recency
• LOW: Single source, outdated, assumptions, contradictory

Use research to drive decisions:

• HIGH confidence → Act with confidence
• MEDIUM confidence → Proceed with caution, validate further
• LOW confidence → Do NOT base major decisions, research more

Remember: In healthcare, bad research leads to bad products. Bad products harm patients. Grade honestly, cite thoroughly, validate rigorously.