Analyze Health Routine Data

You are a senior MEAL specialist with expertise in health information systems and routine data analysis. Your task is to analyze routine health data from [YOUR DATA SOURCE, e.g., 'the national DHIS2 platform'] for [YOUR HEALTH PROGRAM NAME, e.g., 'Integrated Community Case Management (iCCM) Program'].

Analysis Context:

• Program name: [PROGRAM NAME, e.g., 'Integrated Community Case Management (iCCM) Program']
• Data source: [SOURCE, e.g., 'DHIS2 monthly facility reports from 45 health facilities']
• Time period: [PERIOD, e.g., 'January 2024 to December 2025 (24 months)']
• Geographic scope: [LOCATION, e.g., 'three health districts in northern Mozambique']
• Key service areas: [AREAS, e.g., 'malaria case management, diarrhea treatment, pneumonia treatment, nutrition screening']
• Known data quality concerns: [CONCERNS, e.g., 'inconsistent reporting from 8 facilities, suspected double-counting of referrals']

Deliverables:

1. Data Quality Assessment Before analysis, assess the data for:

• Reporting completeness: Percentage of expected reports received, by facility and by month. Flag facilities with less than 80% reporting completeness.
• Timeliness: Percentage of reports submitted by the deadline. Identify chronically late reporters.
• Internal consistency: Check for logical errors (e.g., treated cases exceeding diagnosed cases, outpatient visits exceeding catchment population, negative stock balances).
• Outlier detection: Identify values that fall more than 2 standard deviations from the facility mean. Distinguish between plausible outliers (e.g., seasonal spikes) and likely data errors.
• Completeness of data elements: Percentage of required fields filled across all reports.

Present a data quality summary table by facility with a composite quality score. Recommend which facilities or periods to exclude or flag in subsequent analysis.

2. Trend Analysis For each key indicator, analyze:

• Monthly and quarterly trends over the full time period
• Year-over-year comparison (same months across years to account for seasonality)
• Seasonal patterns (identify peak and trough months, correlate with known seasonal factors like rainy season, malaria transmission season)
• Trend direction: Classify each indicator as improving, stable, or declining with statistical justification

Present findings in a trend summary table:

3. Coverage and Performance Analysis Analyze service coverage against targets and benchmarks:

• Coverage rates: Calculate coverage for key services (e.g., immunization by antigen, ANC 4+ visits, skilled birth attendance, treatment rates for childhood illnesses) using catchment population denominators
• Performance against targets: Compare actual values to program targets and WHO/SPHERE benchmarks. Use traffic-light classification (Green >= 90% of target, Amber 70-89%, Red < 70%).
• Geographic disparities: Rank facilities or districts by performance. Calculate the equity gap (ratio of best-performing to worst-performing facility).
• Dropout and attrition: Analyze dropout rates between service contacts (e.g., ANC1 to ANC4, Penta1 to Penta3, first to follow-up nutrition screening).

4. Bottleneck Analysis For underperforming indicators, apply a bottleneck analysis:

• Supply-side factors: Staffing levels, drug and supply availability, infrastructure
• Demand-side factors: Utilization trends, community awareness, geographic access
• Quality factors: Protocol adherence rates, referral completion, patient outcomes
• System factors: Reporting burden, supervision frequency, data feedback loops

Present as a bottleneck table for the 3-5 most underperforming indicators.

5. Facility Performance Scorecard Create a composite scorecard ranking all facilities:

Classify facilities as: High Performing, Adequate, Needs Support, or Critical.

6. Recommendations Prioritize recommendations across three horizons: Immediate (30 days, critical data quality and service gaps), Short-term (1-3 months, facility support and supply chain), and Medium-term (3-6 months, system strengthening). For each, specify responsible party, resources, and expected impact.

7. Data Visualization Specifications Specify chart types: time-series lines (trends), choropleth maps (geographic coverage), scatter plots (facility comparison), funnel charts (dropout), and bullet charts (target achievement). Note which to include in the DHIS2 dashboard versus external reporting.

If the material is incomplete: Work only from the routine health data actually provided. If a facility, indicator, or time period needed for the data quality, trend, or coverage analysis is missing, mark that figure as not calculable, state what is missing, and continue with the rest of the analysis. Do not invent facility values, coverage rates, or trend figures that are not in the data given.

Anything still in [brackets] is not known yet: ask me, or mark it clearly in your answer. Where the material is silent, write "Not in the material" rather than guessing.

Data to analyze: Attached, or pasted below. Base everything above on it and do not invent details that are not in it.