← Back to Research Overview
PRELIMINARY EXAMINATION

Applied Global Health: Burden of Disease & Climate-Health Analysis

PhD Preliminary Exam — Styvers Kathuni
4 Parts · 60 Minutes Each · 4 Hours Total · Open-Resource with AI

Exam Overview

This preliminary examination tests your ability to apply the Global Burden of Disease (GBD) framework, climate-health analysis, and program evaluation methods to a real-world WASH intervention in Kenya's arid and semi-arid lands. The exam is designed around your doctoral research on the DRIP FUNDI program and requires you to synthesize evidence from global health datasets, climate science, and programmatic data.

Total Duration
4 hrs
60 min per part
Parts
4
Sequential completion
Format
Open
All resources + AI
Focus Region
Kenya
ASAL counties

Exam Structure

Part I — Burden of Disease Analysis (60 min)

Use the GBD Results Tool to analyze Kenya's diarrheal disease burden, disaggregated by age, sex, and risk factor attribution.

Part II — Climate-Health Nexus (60 min)

Analyze climate variability and waterborne disease linkages in northern Kenya, including IPCC projections for East Africa.

Part III — Program Effectiveness Framework (60 min)

Design a DALYs-based evaluation of DRIP FUNDI with cost-effectiveness estimation using real program parameters.

Part IV — Policy Synthesis (60 min)

Write a policy brief for Kenya's Ministry of Water synthesizing burden, climate, and effectiveness findings.

General Instructions

Before You Begin

  • Complete each part within 60 minutes; manage your time across all tasks within each part
  • You may use any publicly available resource, including academic databases, data tools, and AI assistants
  • All AI use must be documented with transparency (see AI Policy below)
  • Responses should demonstrate your own analytical reasoning, not merely reproduce AI outputs
  • When citing data, always include the source, year, and any relevant parameters (location, age group, metric)
  • Your final submission is a single-page website deployed to Cloudflare Pages (see Submission Format below)

Submission Format

Submit as a Claude-Generated pages.dev Site

Your exam submission must be a single HTML page deployed to Cloudflare Pages using Claude Code. This is itself a test of your ability to use AI tools for professional knowledge communication.

  • Use Claude Code to generate a well-structured, visually clear HTML page containing all four parts of your exam response
  • Deploy to Cloudflare Pages — your final deliverable is a live pages.dev URL
  • The site should include: a clear title and navigation, all four parts with headings, any tables/figures embedded directly, and your AI Use Statements
  • Design quality is not graded, but the site should be readable, logically organized, and professionally presented
  • Submit the live URL to the exam proctor by the end of the 4-hour window

AI Use Policy

Claude AI Is Approved for This Exam

You are explicitly permitted and encouraged to use Claude (or other AI assistants) as a research and analytical tool throughout this exam. The ability to effectively use AI tools is itself a competency being assessed.

  • Permitted uses: Data interpretation, literature synthesis, computational assistance, drafting support, methodology review, framework development
  • Required transparency: For each part, include a brief "AI Use Statement" (2–3 sentences) describing how you used AI and what you verified independently
  • Critical evaluation required: You must critically evaluate all AI outputs. Do not accept AI-generated statistics, citations, or factual claims without verification against primary sources
  • Your analysis must dominate: AI should augment your expertise, not replace it. Examiners will assess the depth of your own reasoning, not the sophistication of your prompts

What Distinguishes Excellent AI Use

Strong AI Use

Uses AI to accelerate data extraction and literature review; independently verifies all statistics against GBD or primary sources; applies domain expertise to contextualize AI outputs for ASAL settings; AI Use Statement clearly describes what was delegated vs. what was the candidate's own analysis.

Weak AI Use

Copies AI-generated text without critical evaluation; accepts fabricated statistics or citations; fails to contextualize generic AI outputs for the specific Kenya ASAL context; AI Use Statement is vague or absent; analytical reasoning is shallow or entirely AI-derived.

Data Sources & Tools

The following data sources are central to this exam. You should use them directly to extract and verify data for your analyses.

GBD Results Tool

IHME Global Burden of Disease results for DALYs, deaths, incidence, prevalence by cause, location, age, sex, and year.

vizhub.healthdata.org/gbd-results →

GBD Compare

Interactive visualization tool for comparing burden of disease across locations, causes, and risk factors.

vizhub.healthdata.org/gbd-compare →

CHIRPS Rainfall Data

Climate Hazards Group InfraRed Precipitation with Station data. High-resolution rainfall estimates for Africa.

chc.ucsb.edu/data/chirps →

IPCC AR6 WG II

Chapter 9: Africa. Climate change impacts, adaptation, and vulnerability projections for the continent.

ipcc.ch/report/ar6/wg2/chapter/chapter-9 →

WHO/UNICEF JMP

Joint Monitoring Programme for Water Supply, Sanitation and Hygiene. WASH coverage data by country.

washdata.org →

Kenya DHS

Demographic and Health Survey data for Kenya. Subnational indicators for water, sanitation, health outcomes.

dhsprogram.com/Kenya →

Supplementary Sources

You may also draw on

  • Kenya National Bureau of Statistics (KNBS) — County-level census and survey data
  • FEWS NET — Famine Early Warning Systems Network, seasonal food security and climate outlooks
  • NDMA Kenya — National Drought Management Authority, drought bulletins and early warning data
  • GBD Risk Factor studies — Peer-reviewed publications attributing diarrheal disease to WASH risk factors
  • DRIP FUNDI evaluation reports — Baseline (Feb 2024) and Endline (May 2025) available on this site

Part I — Burden of Disease Analysis

PART I 60 MINUTES

Using the IHME Global Burden of Disease Results Tool and related resources, conduct a comprehensive analysis of Kenya's diarrheal disease burden. Your analysis should contextualize the national picture within the ASAL counties where DRIP FUNDI operates.

Tasks

  1. Extract and present DALYs, deaths, and incidence rates for diarrheal diseases in Kenya (2000–2021). Compare Kenya's burden with the sub-Saharan Africa regional average and the global average. Present trends over time and identify any inflection points. Specify the GBD cause hierarchy level you are using (e.g., Level 3: Diarrheal diseases).
  2. Perform risk factor attribution analysis. Using GBD risk factor data, quantify the proportion of Kenya's diarrheal disease DALYs attributable to: (i) unsafe water source, (ii) unsafe sanitation, and (iii) no access to handwashing facility. Compare these attributable fractions across the three geographic levels (Kenya, SSA, Global). Discuss what the relative magnitudes tell you about Kenya-specific intervention priorities.
  3. Disaggregate by age and sex. Present age-specific DALY rates for diarrheal diseases in Kenya, with particular focus on the under-5 population. Calculate the proportion of total diarrheal DALYs borne by children under 5. Discuss any sex differences in burden and what might explain them.
  4. Contextualize for ASAL counties. Using DHS data, Kenya census data, or other subnational sources, discuss how the national GBD estimates likely over- or underestimate the diarrheal disease burden in the five DRIP FUNDI counties (Turkana, Marsabit, Isiolo, Wajir, Garissa). Consider WASH coverage differentials, under-5 mortality rates, and healthcare access as proxy indicators.

Deliverable (as a section on your submitted pages.dev site)

  • A structured analytical section (approximately 1,500–2,000 words) with at least one data table and one figure/visualization rendered in HTML
  • All data must include source, year, and metric specifications
  • AI Use Statement (2–3 sentences)

Part II — Climate-Health Nexus

PART II 60 MINUTES

Analyze the relationship between climate variability and waterborne disease in northern Kenya, drawing on historical climate data and future projections. Your analysis should connect directly to the operational context of borehole-dependent communities in ASAL counties.

Tasks

  1. Characterize historical rainfall patterns and drought events in the DRIP FUNDI counties. Using CHIRPS data, FEWS NET reports, or NDMA drought bulletins, describe rainfall variability and identify major drought events affecting northern Kenya over the past two decades. What was the frequency and severity of drought events in the 2000–2023 period?
  2. Map the causal mechanisms linking drought to diarrheal disease. Construct a conceptual framework (described in text or as a diagram) showing the pathways through which drought and climate variability affect waterborne disease risk in borehole-dependent communities. Include at least: (i) direct effects on water quantity and borehole yield, (ii) effects on water quality through concentration of contaminants, (iii) behavioral pathways (alternative source use, reduced hygiene), and (iv) nutritional pathways (drought → food insecurity → immune suppression → disease susceptibility).
  3. Analyze the seasonality of borehole failures and disease peaks. Using available evidence from the DRIP FUNDI evaluations and published literature, discuss whether there is a temporal relationship between dry seasons/drought events, borehole breakdowns, and peaks in diarrheal disease incidence. What is the expected lag time between water infrastructure failure and disease outcomes?
  4. Assess future climate projections for East Africa. Using IPCC AR6 (particularly Chapter 9: Africa), summarize the projected changes in rainfall patterns, drought frequency, and temperature for the Horn of Africa under SSP2-4.5 (moderate) and SSP5-8.5 (high emissions) scenarios. What do these projections imply for the future demand on borehole infrastructure and the likely trajectory of water-related disease burden?

Deliverable (as a section on your submitted pages.dev site)

  • A structured analytical section (approximately 1,500–2,000 words) including a causal framework rendered as an HTML diagram or structured text
  • Integration of at least 3 distinct data sources (e.g., CHIRPS, IPCC, GBD, DHS, NDMA)
  • AI Use Statement (2–3 sentences)

Part III — Program Effectiveness Framework

PART III 60 MINUTES

Design a DALYs-averted evaluation framework for the DRIP FUNDI program. This part requires you to combine GBD methodology with real program parameters to estimate the potential health impact of improved borehole uptime.

Tasks

  1. Review the DALY methodology. Briefly explain the construction of the DALY metric, including Years of Life Lost (YLL), Years Lived with Disability (YLD), and disability weights for diarrheal diseases. What are the current GBD disability weights for mild, moderate, and severe diarrheal episodes? What is the standard life expectancy reference used in current GBD calculations?
  2. Construct a causal pathway framework. Map the logical chain from DRIP FUNDI's core intervention (improved borehole uptime through sensor-enabled monitoring and rapid repair) to DALYs averted. Your framework should trace: borehole uptime → population with improved water access → reduced exposure to unsafe water sources → reduced diarrheal disease incidence → DALYs averted. Identify assumptions at each step and where evidence is strong vs. uncertain.
  3. Perform an estimation exercise using real parameters. Using the following inputs, estimate the potential annual DALYs averted by DRIP FUNDI:
    • Target population: 195,000 beneficiaries across 260 boreholes
    • Baseline uptime: 84.4%; Endline uptime: 85.7% (from evaluation data)
    • GBD diarrheal disease DALY rate for Kenya (from your Part I analysis)
    • Risk factor attributable fraction for unsafe water (from your Part I analysis)
    • Assumed relative risk reduction from improved water access (cite your source)
    Show your calculations explicitly. Conduct a sensitivity analysis varying at least two key parameters.
  4. Compare cost-effectiveness with alternative WASH interventions. Using published cost-effectiveness data (e.g., from DCP3, WHO-CHOICE, or peer-reviewed CEA literature), compare the estimated cost per DALY averted for DRIP FUNDI (using the program's $2M budget for Phase 1) with: (i) household water treatment (e.g., chlorination), (ii) community-level water supply improvements, and (iii) sanitation interventions. Where does DRIP FUNDI fall on the cost-effectiveness spectrum?

Deliverable (as a section on your submitted pages.dev site)

  • A structured evaluation framework section with explicit calculations (approximately 2,000–2,500 words)
  • Causal pathway rendered as an HTML diagram or detailed textual framework
  • Sensitivity analysis as an HTML table showing range of DALYs averted estimates
  • Cost-effectiveness comparison as an HTML table
  • AI Use Statement (2–3 sentences)

Part IV — Policy Synthesis

PART IV 60 MINUTES

Drawing on your analyses from Parts I–III, write a policy brief addressed to Kenya's Ministry of Water, Sanitation and Irrigation. The brief should be actionable, evidence-based, and grounded in the specific context of ASAL counties.

Tasks

  1. Synthesize the evidence base. In 1–2 paragraphs, summarize the key findings from your burden of disease analysis (Part I), climate-health nexus assessment (Part II), and program effectiveness framework (Part III). Frame the synthesis around the central policy question: How should Kenya invest in water infrastructure maintenance to reduce the climate-sensitive disease burden in ASAL communities?
  2. Develop 3–4 actionable policy recommendations. Each recommendation should:
    • Be specific and implementable (not vague aspirations)
    • Cite the evidence basis from your earlier analyses
    • Include an estimated scope of impact where possible
    • Consider political and financial feasibility in the Kenyan ASAL context
  3. Assess generalizability. To what extent are your findings and recommendations applicable beyond the five DRIP FUNDI counties? Consider: (i) other ASAL counties in Kenya, (ii) analogous contexts in the Horn of Africa (Ethiopia, Somalia), and (iii) limitations on transferability. What conditions must hold for your recommendations to apply elsewhere?
  4. Identify priority research gaps. Based on the limitations you encountered in Parts I–III, identify 3–4 critical research questions that would strengthen the evidence base for water security investments in climate-stressed regions. For each gap, explain what data or study design would be needed and why it matters for policy.

Deliverable (as a section on your submitted pages.dev site)

  • A policy brief section (approximately 1,500–2,000 words) with clear headings, an executive summary, and recommendations
  • The brief should be written for a ministerial audience — accessible, evidence-based, and action-oriented
  • AI Use Statement (2–3 sentences)

Evaluation Rubric

Your exam will be evaluated on the following weighted criteria. Each part is scored out of 100 points, then weighted according to the table below.

Component Weight Evaluation Criteria
Part I — Burden of Disease 20% Accuracy of GBD data extraction; appropriate use of metrics (DALYs, rates vs. counts); quality of risk factor attribution analysis; strength of subnational contextualization for ASAL counties
Part II — Climate-Health Nexus 20% Quality of climate data synthesis; rigor of causal framework; integration of multiple data sources; appropriate use of IPCC projections; connection to borehole infrastructure
Part III — Effectiveness Framework 30% Soundness of DALY methodology review; logical coherence of causal pathway; transparency and accuracy of calculations; quality of sensitivity analysis; appropriate cost-effectiveness comparison
Part IV — Policy Synthesis 20% Quality of evidence synthesis; specificity and feasibility of recommendations; thoughtful generalizability assessment; identification of meaningful research gaps; appropriate tone for ministerial audience
AI Use & Site Delivery 10% Transparency of AI use documentation; critical evaluation of AI outputs; evidence of independent verification; appropriate delegation vs. own analysis; successful deployment of a well-structured pages.dev site using Claude Code

Scoring Guide

Distinction (85–100)

Demonstrates mastery of GBD methodology and climate-health analysis. Calculations are accurate with well-reasoned assumptions. Policy recommendations are specific, feasible, and strongly grounded in evidence. AI use is transparent, critical, and value-adding.

Pass with Merit (70–84)

Shows solid command of methods and data sources. Analysis is generally sound with minor gaps. Recommendations are evidence-based but may lack specificity. AI use is documented and mostly well-integrated.

Pass (55–69)

Demonstrates adequate understanding of core concepts. Some analytical weaknesses or unsupported assumptions. Recommendations are reasonable but generic. AI use documentation is present but limited.

Below Threshold (<55)

Significant gaps in methodology or data use. Calculations contain major errors. Recommendations lack evidence basis. AI outputs accepted uncritically or use is undocumented.