HPAI Outbreak: General Epidemic Description

Phase 1: December 20, 2025 – January 13, 2026

Published

February 16, 2026

Outbreak Overview

As of January 13, 2026, a total of 103 HPAI-confirmed outbreaks have been reported across 8 counties and 59 districts since the index case on December 22, 2025 (22 days). Both species present in the poultry population (duck, chicken) are affected. Of the 103 confirmed farms, 51 have completed culling, 11 are in progress, and 41 are planned.

Distribution by Species and Production Type

Outbreak counts

Table 1: Confirmed HPAI outbreaks by species and production type, with attack rate relative to total farm population.
Species Production Type Total Farms Outbreaks Attack Rate (%)
Chicken Broiler 1 1248 18 1.4
Chicken Broiler 2 2746 56 2.0
Chicken Layer 441 7 1.6
Duck Conventional 3691 11 0.3
Duck Organic 1034 11 1.1
Chicken All 4435 81 1.8
Duck All 4725 22 0.5
Total All 9160 103 1.1

Capacity affected

Table 2: Estimated animal capacity affected by species and production type.
Species Production Type Outbreaks Total Capacity Mean Capacity / Farm
Chicken Broiler 1 18 89,916 4,995
Chicken Broiler 2 56 122,505 2,188
Chicken Layer 7 125,448 17,921
Duck Conventional 11 168,927 15,357
Duck Organic 11 9,381 853
Total All 103 516,177 5,011

Detection method

Table 3: Outbreaks by detection method and species.
Species Contact Tracing Passive Preshipment
Chicken 1 74 6
Duck 0 22 0
Total 1 96 6

Epidemic Timeline

Daily epidemic curve

Figure 1: Daily confirmed HPAI outbreaks (bars) with 7-day rolling average (red line). Bars are coloured by species.

Cumulative epidemic curve

Figure 2: Cumulative confirmed HPAI outbreaks over time, coloured by species.

Incidence by epidemiological week

Table 4: Weekly confirmed outbreaks by species and cumulative total.
Week Period Duck Chicken Week Total Cumulative
1 Dec 22 – Dec 28 0 4 4 4
2 Dec 29 – Jan 04 3 17 20 24
3 Jan 05 – Jan 11 12 36 48 72
4 Jan 12 – Jan 13 7 24 31 103

Spatial Distribution

Outbreak map

Figure 3: Spatial distribution of confirmed HPAI outbreaks (coloured by species) relative to the total farm population (grey points). County boundaries shown.

Outbreaks by county

Table 5: Outbreaks by county, with total farm population and attack rate.
County Total Farms Outbreaks Attack Rate (%)
Berks 1151 44 3.8
Susquehanna 572 22 3.8
Indiana 515 20 3.9
Allegheny 820 5 0.6
Luzerne 663 4 0.6
Lancaster 654 3 0.5
Lycoming 469 3 0.6
Cumberland 854 2 0.2

Temporal-spatial progression

Figure 4: Spatial progression of outbreaks over time. Point colour indicates timing (darker = earlier). County boundaries shown for reference.

Production type map

Figure 5: Spatial distribution of outbreaks coloured by production type.

Movement Network

The movement dataset records 7,187 animal movements between 3,994 farms from September 01, 2025 to January 13, 2026. The network has a directed structure: 1,248 source farms (all broiler_1) ship to 2,746 destination farms (all broiler_2). This represents a broiler supply chain. Duck farms (conventional and organic) are entirely absent from the movement records.

Network summary

Table 6: Movement network summary statistics.
Metric Value
Total movements recorded 7,187
Date range Sep 01, 2025 – Jan 13, 2026
Unique source farms (all broiler 1) 1,248
Unique destination farms (all broiler 2) 2,746
Total farms in network 3,994
Case farms in network 74 of 103
— as source (broiler 1) 18
— as destination (broiler 2) 56

Movement volume over time

Figure 6: Weekly movement volume. The vertical dashed line marks the date of the first confirmed outbreak (December 22, 2025). Movement volumes decline during the outbreak period.

Case farm involvement

Table 7: Movements involving confirmed case farms: timing relative to confirmation and direct case-to-case links.
Metric Count
Movements FROM case farms 88
— before confirmation 85
— on day of confirmation 3
— after confirmation 0
Movements TO case farms 165
— before confirmation 165
Direct case → case movements 3

Network visualisation

Figure 7: Movement network graph. Nodes represent farms; edges represent recorded movements (aggregated). Node colour indicates case status; node shape indicates production type (broiler 1 = circle, broiler 2 = triangle). Only the connected component containing at least one case farm is shown to reduce visual clutter.

Culling Response

Confirmed case culling

Table 8: Culling status for confirmed HPAI outbreaks and preventive culls.
Type Completed In Progress Planned Total
Confirmed outbreaks 51 11 41 103
Preventive culls 11 1 40 52
Figure 8: Timeline of culling activity for confirmed outbreaks. Each bar represents one farm, ordered by confirmation date. Colour indicates cull status. The gap between confirmation (left edge) and cull start shows the response delay.

Time from confirmation to culling

Table 9: Time from confirmation to cull start (days) for confirmed outbreaks with culling data.
N Mean Median Min Max
103 2.9 3 1 6

Preventive culling

A total of 52 preventive culls have been ordered on farms not (yet) confirmed positive. These target farms identified through proximity or epidemiological links to confirmed cases.

Table 10: Preventive culls by species, production type, and status.
Species Production Type Planned Completed In Progress Total
Duck Conventional 20 4 1 25
Chicken Broiler 2 7 3 0 10
Chicken Broiler 1 8 0 0 8
Duck Organic 4 4 0 8
Chicken Layer 1 0 0 1

Questions arising from the data

The descriptive patterns above raise several questions for further investigation:

  • Is the epidemic accelerating, stabilising, or decelerating? The weekly incidence table and epidemic curve provide a basis for this assessment. Growth rate estimation (e.g., doubling time, effective reproduction number) would quantify the trajectory more precisely — see the companion temporal analysis for these estimates.

  • Why are chicken farms, particularly broiler type 2, disproportionately affected? Attack rates differ substantially across species and production types. Possible explanations include spatial confounding (chicken farms may be concentrated near the outbreak focus), differences in biosecurity practices by production type, differences in susceptibility or clinical presentation affecting detection probability, or transmission via production-specific supply chains. Disentangling these requires spatial risk factor analysis.

  • Does the broiler supply chain explain the concentration of cases in broiler 2 farms? The movement network is a directed broiler 1 → broiler 2 supply chain. Broiler 2 farms have the highest attack rate and receive animals from broiler 1 farms, some of which are also confirmed cases. However, only 3 direct case-to-case movements were recorded. Further analysis could quantify whether farms with more incoming movements, or movements from case farms specifically, had higher infection risk — controlling for spatial proximity.

  • Is passive surveillance sufficient? The overwhelming majority of detections are through passive surveillance (clinical suspicion). The low contribution of preshipment testing and contact tracing raises the question of whether subclinical or early-stage infections are being missed, and whether the true extent of the epidemic is larger than confirmed cases suggest.

  • Are culling operations keeping pace with new detections? The cull status data show a growing backlog of planned and in-progress culls. Whether delays in depopulation are contributing to onward transmission is a question for operational and transmission modelling.

  • Where is the epidemic likely to spread next? The temporal-spatial progression suggests outward expansion from an initial focus. Identifying the leading edge and farms at highest risk of future infection would inform targeted surveillance and preventive measures.

Limitations and analytical choices

Data not included in this report

  • Mortality ledger data: Individual farm mortality records (PDF format) are available for a subset of farms. These could provide insight into within-farm dynamics and time from infection to detection, but were not incorporated here.

  • Farm activity data: Production activity logs (activity.csv) could contextualise which farms were actively stocked at the time of the outbreak, but were not used in this descriptive analysis.

  • High-risk zone and wetland layers: Geospatial data on the designated high-risk zone (hrz_32626.geojson) and wetland/water body coverage (clc_32626.geojson) are available and could add context to the spatial maps, but were omitted to keep the maps focused on outbreak distribution.

Analytical limitations

  • Attack rates are crude: The attack rates presented use total registered farms as the denominator. Not all farms may have been actively stocked during the outbreak period, so true attack rates among at-risk farms may differ.

  • No adjustment for spatial confounding: Differences in attack rates across species and production types may be partly or wholly explained by the geographic concentration of the outbreak rather than true biological differences.

  • Epidemic curve right-truncation: The most recent days of the epidemic curve are likely incomplete due to delays between suspicion and confirmation. The apparent case count in the final days may underestimate the true incidence.

  • Movement network is descriptive only: The movement section presents network structure and case involvement but does not model whether movements are associated with infection risk. Temporal precedence (movements before confirmation) does not establish causation. The network also only covers chicken broiler farms, so transmission pathways for duck and layer farms remain unobserved.

  • No formal statistical testing: This report is purely descriptive. Differences in attack rates have not been tested for statistical significance, and no regression or spatial modelling has been applied.