SRS / Piscirickettsiosis — Decision Support Prototype v2

Farm-level SRS risk & intervention calculator

Weighted score from environmental, spatial, parasitic and management risk factors, now paired with an evidence-ranked intervention checklist and antibiotic-timing guidance. Weights are literature-derived starting points — calibrate against your own mortality and treatment records before operational use.

Farm & production context

Species and cycle stage set the baseline hazard curve — rainbow trout has the shortest median time to first outbreak, coho the longest.
Median time to first SRS outbreak: Atlantic ~48 wk, rainbow trout ~33 wk (highest hazard), coho >50% of cycles never outbreak.
SRS mortality risk is consistently highest in summer, ~10x the winter median.
Retrospective cohort of 571 cycles (Gaete-Carrasco et al. 2026): median time to outbreak — spring 47w (ref., latest), summer 42w, winter 38w, autumn 35w (earliest). Only autumn differs significantly from spring (HR=1.15, p=0.029); winter and summer are not statistically distinguishable from spring.
Weekly SRS-attributed mortality at which antibiotic treatment was first initiated. Delaying initiation to ≥0.03% is associated with significantly earlier onset of the next outbreak (HR=1.37, p<0.0001, n=571 cycles) compared with initiating at <0.01% — the strongest single predictor identified in the Weibull model.

Spatial / neighborhood risk

Between-farm connectivity is consistently the strongest predictor of SRS across studies.
Expert-elicited risk threshold ≈2 active farms in a 5 km ring (ROC-derived cut-off, AUC 0.82).

Sea lice (Caligus) & co-infection

The single strongest actionable predictor of SRS onset in the largest retrospective cohort to date.
>1 egg-laying female/fish = "high-load" report; strongest predictor of both weekly mortality risk and hazard of first outbreak.
Counterintuitively risk-increasing above ~2 treatments — likely treatment-associated stress rather than the parasite itself.

Husbandry & biosecurity stressors

Strongest single predictor in weekly mortality models: weeks after an initial outbreak carry markedly higher risk than the initial outbreak itself.

Current risk

0
—
0LowModerateHigh100
—
Adjust inputs to see recommended actions.
Antibiotic-timing note
—

Projected cumulative outbreak probability — this production cycle

Curve shape reflects species-specific median time-to-outbreak, then shifted earlier by neighborhood exposure and Caligus load. Illustrative — not a fitted survival model.

Evidence-ranked control measures

From a structured elicitation of 9 head fish-health managers (covering ~59% of Chilean Atlantic salmon and ~65% of rainbow trout production) scoring 67 SRS control measures for effectiveness and consensus. Tier 1 = mean effectiveness ≥3.5/4 with high inter-expert consensus; Tier 2 = mean ≥3.0 with moderate consensus. Toggle measures your farm already has in place to see the residual risk.
TIER 1High priority — high effectiveness, high consensus
TIER 2Medium priority — good effectiveness, moderate consensus
Note: therapeutic (antibiotic) and vaccination measures were rated lower and with far greater inter-expert variability than the measures above — vaccines are understood to delay rather than prevent outbreaks, and antimicrobial effectiveness depends heavily on timing of initiation (see antibiotic-timing note on the calculator tab).

Residual risk after interventions

—
Starting score: — → after checked measures
Mitigation weights are illustrative placeholders reflecting relative rank from the expert-elicitation study, not measured absolute risk reductions — calibrate against farm outcome data when available.

What's built vs. what's next

This prototype is the calculator layer. Two extensions are designed for but not yet implemented:
①
Farm-level statistical disease model
Replace the illustrative weighted score with a fitted mixed-effects negative binomial model (weekly SRS mortality) and Cox survival model (time-to-first-outbreak), following the two-model structure used on the 2014–2021 Chilean SIFA dataset (652 farms, 1,546 production cycles). Needs farm-week-level records: mortality by cause, Caligus counts, treatment dates, species, geography.
②
Integrated sea lice (Caligus) management module
A linked scheduling module treating Caligus control as an SRS-prevention action rather than a separate parasite program — since female egg-laying lice load is the strongest actionable predictor of SRS onset, and bath treatments themselves carry a treatment-stress risk above ~2 per cycle. Would output a delousing-timing recommendation jointly optimized against SRS hazard, not lice count alone.

Evidence-strength legend (from the field-evidence systematic review)

70 field-based studies (1997–2025) synthesized across a five-level prevention framework. Strength ratings below carry through to the factors and measures used in this calculator.
● Strong — ≥3 independent field studies, consistent direction ◉ Moderate — 1–2 field studies, or inconsistent ○ Weak / indirect — modelling or lab-only evidence
Prevention levelSelected factorStrength
PrimordialHydrodynamic connectivity / upstream infected farms●
PrimordialWater temperature & salinity●
PrimaryCaligus infestation → increased SRS mortality risk●
PrimaryVaccination delays but does not prevent outbreaks◉ (only 2 field-effectiveness studies of >30 licensed products)
PrimaryFallowing ≥40–90 days reduces early-cycle outbreak probability●
SecondaryActive/risk-based surveillance → earlier detection, lower post-treatment mortality●
TertiaryEarly antibiotic initiation vs. late/high-threshold initiation● (linked to earlier outbreak recurrence when delayed)
Tertiary~40% of florfenicol-treated fish fail to reach therapeutic tissue concentration under field conditions◉
QuaternarySub-inhibitory antibiotic exposure → AMR gene selection & biofilm induction◉
Prototype for illustration only. Weights, mitigation values, and curves are derived from published epidemiological and expert-elicitation literature on Piscirickettsia salmonis (SRS) in farmed salmonids and are not calibrated to any specific farm, region, or dataset. This tool does not replace veterinary diagnosis — confirm suspected outbreaks with PCR/histopathology before treatment decisions. Antibiotic use should follow local veterinary prescription and stewardship guidelines.
Sources referenced in this version: Estévez et al. 2019, Aquaculture 507 (expert-elicited risk/protective factors, RR/ARR). Gaete-Carrasco et al. 2026, Aquaculture 625:744273 (retrospective cohort of 571 Atlantic salmon cycles, 2013–2019; Weibull AFT model of time-to-first-outbreak; stocking season and antibiotic-initiation threshold as significant predictors). Diethelm-Varela et al. 2025, Journal of Fish Diseases e70097 (SIFA retrospective cohort 2014–2021, sea lice & SRS mortality risk). Bustos, Diethelm-Varela, Riofrio, Mancilla & Mardones 2026 (expert prioritization of 67 SRS control measures). Diethelm-Varela et al., field-based systematic review & Swiss Cheese evidence map of P. salmonis management (1997–2025).