---
license: mit
language:
- en
pretty_name: ABX-CS-008 Pediatric Dosing Effectiveness Decay
tags:
- antibiotics
- clinical
- pediatrics
- dosing
- weight-based
- pk-pd
- tabular
task_categories:
- tabular-classification
size_categories:
- n<1k
---

ABX-CS-008 Pediatric Dosing Effectiveness Decay

Purpose

Detect early pediatric dosing decoupling when weight-based dosing no longer predicts exposure and response.

Core pattern

- severity_index high
- dose_mg_per_kg_per_day is near expected
- pediatric_coherence_index drops
- exposure_shortfall_vs_expected stays high
- efficacy_shortfall_vs_expected stays high
- symptom_nonresponse_vs_expected stays high
- later_dose_escalation_or_abx_change_or_prolonged_admission_flag appears

Files

- data/train.csv
- data/test.csv
- scorer.py

Schema

Each row is one timepoint in a within series pediatric antibiotic course.

Required columns

- row_id
- series_id
- day
- host_model
- age_years
- weight_kg
- abx_drug
- dose_mg_per_kg_per_day
- expected_dose_mg_per_kg_per_day
- dose_deviation_mg_kg
- exposure_index
- expected_exposure_index
- exposure_shortfall_vs_expected
- efficacy_index
- expected_efficacy_index
- efficacy_shortfall_vs_expected
- symptom_score
- expected_symptom_score
- symptom_nonresponse_vs_expected
- severity_index
- pediatric_coherence_index
- later_dose_escalation_or_abx_change_or_prolonged_admission_flag
- assay_method
- source_type
- pediatric_dosing_effectiveness_decay_signal
- earliest_pediatric_dosing_effectiveness_decay

Labels

- pediatric_dosing_effectiveness_decay_signal
  - 1 for rows at or after first confirmed decay onset

- earliest_pediatric_dosing_effectiveness_decay
  - 1 only for the first onset row in that series

Scorer logic in v1

- require severity_index at least 0.80
- require actual dose at least 90 percent of expected dose
- candidate onset point
  - pediatric_coherence_index at most 0.40
  - exposure_shortfall_vs_expected at least 0.25
  - efficacy_shortfall_vs_expected at least 0.20
  - symptom_nonresponse_vs_expected at least 2.0
  - for two consecutive points
  - ignore one point exposure shortfall spike then recovery
- confirmation
  - later_dose_escalation_or_abx_change_or_prolonged_admission_flag equals 1 later in series

Evaluation

Run

- python scorer.py --path data/test.csv