Estimating the Perinatal Health Benefits of Hypothetical Pollution Interventions in Santiago, Chile Using Parametric G-Computation 🏭 👶

💰 Funding

FONDECYT Nº 11240322: Climate change and urban health: how air pollution, temperature, and city structure relate to preterm birth

Additional support: (CR)², Chile, FONDAP/ANID 1523A0002

👥 Research Team

📬 Estela Blanco (estela.blanco@uc.cl), Principal Investigator / Corresponding Author

📬 José Daniel Conejeros (jdconejeros@uc.cl), Research Assistant / Repository Manager

Research Collaborators: Ismael Bravo, Felipe Cornejo, Axel Osses, & Tarik Benmarhnia

📌 Publication

Work in progress.

---

🎯 Project Overview

Background

Ambient air pollution is a major environmental risk factor for adverse perinatal outcomes. Preterm birth (delivery before 37 completed weeks of gestation) remains a leading cause of neonatal mortality and long-term morbidity. While epidemiological evidence links PM₂.₅, NO₂, and O₃ to preterm birth, fewer studies have translated observed exposure–response associations into population-level estimates of the health benefits of realistic pollution-reduction scenarios, particularly in Latin American urban settings with heterogeneous exposure patterns.

Objective

To estimate the population-level impact of hypothetical reductions in PM₂.₅, NO₂, and O₃ on the cumulative risk of preterm birth among singleton births in urban Santiago, Chile (2010–2020), using distributed-lag Cox models combined with parametric g-computation.

Methods (as reported in the manuscript)

• Design: Retrospective population-based cohort (DEIS birth records, 2010–2020).
• Study area: Urban conurbation of Santiago (32 municipalities in the Province of Santiago plus Puente Alto; 33 municipalities total).
• Analytic sample: 713,918 singleton live births with gestational age ≥28 weeks at delivery, complete covariates, plausible birthweight-for-gestational age (Alexander et al., 1996), and no fixed-cohort bias (gestation start ≥ January 1, 2010; delivery on or before March 20, 2020). 51,081 preterm births (7.2%). Exclusions in Figure 1.
• Exposure: Daily PM₂.₅, NO₂, and O₃ from the national monitoring network; municipality-day concentrations via ordinary kriging to municipal administrative centers (primary). IDW in parallel for sensitivity. Values below detection limits set to the limit. Weekly series (X_{iw}) (weeks 1–44) and weekly mean temperature; full-pregnancy NDVI (MODIS MOD13Q1, Kalman-imputed daily values); municipal SOVI (low, medium-low, medium-high).
• Outcome: Preterm birth (<37 completed weeks).
• Distributed-lag models (main text): For each gestational week (w = 1,\ldots,44), separate Cox models on completed gestational age with delayed entry at week 28, Efron ties, time-weighted lag through week (w-1) (equation 1 in the manuscript), and multipollutant adjustment (concurrent weekly co-pollutants plus weekly and lagged terms for the target pollutant). Covariates: newborn sex; parental age, education, occupation; conception month and year; COVID-19 indicator (deliveries after March 1, 2020); SOVI; weekly temperature; full-pregnancy NDVI. Single-pollutant DLMs and trimester/overall averages are in the supplement (Tables S5–S10, Figure S6).
• G-computation (main text): Parametric g-formula using the multipollutant natural-course Cox models (risk weeks 28–36). Counterfactual histories apply a 20% proportional reduction ((\pi = 0.20)) to all gestational weeks of the target pollutant, with lag terms recomputed; co-pollutants, temperature, and covariates stay at observed values. Discrete-time risks from Breslow baseline hazards; global effects at week 36 (prevalence, expected cases, RR, RD, AR, PAF). Single-week 20% interventions for critical-window heatmaps (Figure 5). 95% CIs: 2.5th–97.5th percentiles of a parametric bootstrap (250 replicates) resampling Cox coefficients (supplementary methods S1, Table S1).
• Supplementary g-computation: 10%, 20%, and 30% proportional reductions under single-pollutant weekly Cox models (not reported in the main tables).

Key findings (main text)

• Exposure (2010–2020): Mean daily kriging concentrations about 25.9 µg/m³ (PM₂.₅), 21.3 ppbv (NO₂), and 14.0 ppbv (O₃); strong winter–summer contrasts (Figure 2; Table S2). Kriging and IDW agreed closely (e.g., (r = 0.96) for PM₂.₅; Figure S3).
• DLM (Figure 3): Multipollutant week-specific HRs per 1-unit increase were near the null but differed by timing (e.g., O₃ elevated in early pregnancy and week 34; NO₂ inverse in early gestation and positive around weeks 25–32; PM₂.₅ small positive associations in selected early weeks).
• G-formula, 20% reduction (Table 2): Natural-course cumulative risk at week 36 6.84% (95% CI 5.43 to 8.90), 48,863 expected preterm births.
  • O₃: 6.58%; RR 0.961 (0.945 to 0.977); RD −0.27 pp (−0.44 to −0.14); PAF 3.88% (2.32 to 5.60); about 1,900 fewer preterm births.
  • NO₂: 6.96%; RR 1.017 (1.001 to 1.030); RD +0.11 pp (0.01 to 0.21); PAF −1.67% (−3.00 to −0.07).
  • PM₂.₅: 6.85%; RR 1.001 (0.988 to 1.010); RD +0.01 pp (−0.09 to 0.07); PAF −0.08% (−0.98 to 1.21).
• Timing (Figure 5): Single-week 20% O₃ reductions lowered cumulative risk for all intervention weeks, largest late in pregnancy (e.g., week 34: −0.046 pp at week 36). NO₂ single-week reductions mostly increased risk slightly; PM₂.₅ risk differences were near zero.

---

ℹ️ Additional analyses in this repository (not in the main manuscript)

The repository implements the full data pipeline and many secondary and sensitivity analyses that support the supplement or internal robustness checks. They are not summarized in the main-text tables or figures above.

Topic	Scripts (examples)	Outputs (examples)
Single-pollutant weekly DLM	9.0, 9.1	02_Output/Models/DLM_models_krg.png, DLM_cox_*
Multipollutant weekly DLM (manuscript Figure 3)	13.0, 13.1	DLM_multi_models_krg.png, DLM_multi_cox_*
Period-average Cox (trimester / full pregnancy)	11.x, 15.x	02_Output/Exposure_PO/, Exposure_PO_Multi/
G-computation, 17 scenarios (caps + 10/20/30%), single-pollutant weekly models	10.0–10.4	02_Output/G-Form/
G-computation, 20% multipollutant (manuscript Table 2, Figures 4–5)	14.0–14.4	02_Output/G-Form-Multi/
G-computation on trimester- or full-pregnancy exposure aggregates	12.x, 16.x	02_Output/G-Form-Period/, G-Form-Period-Multi/
Exposure descriptives beyond Figure 2 (maps, time series, annual tables)	7.0, 7.1, 8.0	02_Output/Descriptives/
Positivity diagnostics by gestational week	7.2	Table_positivity_by_week.xlsx
IDW vs kriging comparisons	All *_idw* outputs	Supplementary Figures S1, S3, S6; Tables S8, S10

Practical note: Default code settings may use 500 bootstrap replicates (GFORM_DEFAULTS$boot_iter in 10.1); the manuscript reports 250. Reproduce published CIs with GFORM_BOOT_ITER=250. Cap-threshold interventions (min(X, c)) and 10%/30% proportional reductions are implemented under 10.x but are not part of the main-text g-computation results (only 20% multipollutant reductions are).

---

Code Structure

Setup Scripts

• 00_Code/0.1 Settings.R: global settings and locale
• 00_Code/0.2 Packages.R: package installation and loading (main pipeline)
• 00_Code/0.2 Packages_gform.R: packages for g-computation pipeline
• 00_Code/0.3 Functions.R: custom helper functions

Data Processing Scripts

• 00_Code/1.0 Pollution_process_data.R: load and clean interpolated PM₂.₅, NO₂, O₃ series
• 00_Code/2.0 Births_process_data.R: birth data cleaning, cohort definition, exclusions
• 00_Code/3.0 NDVI_EarthEngine_commune_extraction.py: NDVI extraction (Google Earth Engine)
• 00_Code/3.1 Temp_NDVI_data.R: temperature and NDVI processing
• 00_Code/4.0 Climate_data_generate.R: climate data generation
• 00_Code/5.0 Exposure_data_births.R: weekly gestational exposure histories
• 00_Code/6.0 Join_full_data.R: merge pollution, climate, and birth data
• 00_Code/8.0 Correlation_pollulants.R: pollutant correlation analysis

Descriptive Analysis

• 00_Code/7.0 Descriptive_births.R: birth and preterm trends
• 00_Code/7.1 Descriptive_exposition.R: exposure descriptives (Table S2 source, Figure 2, repository maps and time series)
• 00_Code/7.2 Positivity_analysis.R: positivity by gestational week

Statistical Models (manuscript and supplement)

• 00_Code/9.0 DLM_pollution.R, 9.1 DLM_plots.R: single-pollutant weekly DLMs (supplement)
• 00_Code/13.0 DLM_multi_pollution.R, 13.1 DLM_multi_plots.R: multipollutant weekly DLMs (Figure 3)
• 00_Code/11.x, 15.x: period-average exposure Cox models (Tables S5–S6)

G-Computation: manuscript (multipollutant, 20%)

• 00_Code/14.0 G-Form_multi_functions.R: multipollutant weekly Cox + g-formula
• 00_Code/14.1 G-Form_multi_build_interventions.R: counterfactual histories (pct20 × 3 pollutants)
• 00_Code/14.2 G-Form_multi_models.R: models, bootstrap, single-week heatmaps
• 00_Code/14.3 G-Form_multi_plots.R: Figures 4–5 (02_Output/G-Form-Multi/Figures/)
• 00_Code/14.4 G-Form_multi_table.R: Table 2 summary exports

G-Computation: extended scenarios (repository only)

• 00_Code/10.0–10.4: 17 global scenarios (caps and 10/20/30% reductions) under single-pollutant weekly models (02_Output/G-Form/)
• 00_Code/12.x, 16.x: g-formula on trimester/full-pregnancy exposure aggregates
• 00_Code/intervention.txt: scenario list reference

G-Computation Intervention Registry (10.x, repository extension)

Scenarios are defined in GFORM_INTERVENTION_REGISTRY (00_Code/10.1 G-Form_functions.R). Stage 1 (10.0) writes one RDS per scenario to 02_Output/G-Form/Interventions/; Stage 2 (10.2) runs models and bootstrap by intervention_number (1–17) or via GFORM_INTERVENTIONS. These 17 scenarios are not reported in the main manuscript (see table for stubs used in sensitivity work and composite figures under G-Form/Figures/).

#	Pollutant	Scenario	Registry ID	Output stub
1	PM₂.₅	−20% all weeks	pm25_krg_pct20	pm25_pct20
2	NO₂	−20% all weeks	no2_krg_pct20	no2_pct20
3	O₃	−20% all weeks	o3_krg_pct20	o3_pct20
4	PM₂.₅	< 20 µg/m³	pm25_krg_lt20	pm25_lt20
5	PM₂.₅	< 5 µg/m³	pm25_krg_lt5	pm25_lt5
6	NO₂	< 20 ppbv	no2_krg_lt20	no2_lt20
7	NO₂	< 5 ppbv	no2_krg_lt5	no2_lt5
8	PM₂.₅	< 15 µg/m³	pm25_krg_lt15	pm25_lt15
9	PM₂.₅	< 10 µg/m³	pm25_krg_lt10	pm25_lt10
10	NO₂	< 15 ppbv	no2_krg_lt15	no2_lt15
11	NO₂	< 10 ppbv	no2_krg_lt10	no2_lt10
12	PM₂.₅	−10% all weeks	pm25_krg_pct10	pm25_pct10
13	PM₂.₅	−30% all weeks	pm25_krg_pct30	pm25_pct30
14	NO₂	−10% all weeks	no2_krg_pct10	no2_pct10
15	NO₂	−30% all weeks	no2_krg_pct30	no2_pct30
16	O₃	−10% all weeks	o3_krg_pct10	o3_pct10
17	O₃	−30% all weeks	o3_krg_pct30	o3_pct30

Cap semantics: < 5 means exposure is fixed at 5 units when the observed weekly value exceeds 5 (min(X, c)), not a subtraction of 5 from observed concentrations.

---

📈 Manuscript Figures and Tables (September 2026)

Figure 1. Flowchart of Analytical Sample Construction (2010–2020)

Starting from 2,557,140 singleton births in Chile (2010–2020), sequential exclusions yielded 713,918 births in urban Santiago (51,081 preterm, 7.2%). See 03_Paper/Gformula_Santiago_PTB_29092026.docx.

Table 1. Cohort Characteristics

Descriptive statistics for the full analytic cohort and preterm births (sex, parental characteristics, SOVI). Generated from the cleaned cohort (2.0); values in the manuscript Table 1.

Figure 2. Distribution of Daily PM₂.₅, NO₂, and O₃ (Kriging)

Note: Municipality-day concentrations from ordinary kriging. IDW analogue: Supplementary Figure S1 (Histogram_IDW_panel_compiled.png).

Figure 3. Hazard Ratios for Preterm Birth by Gestational Week (Multipollutant DLM)

Note: Multipollutant distributed-lag Cox models (kriging); HRs per 1-unit increase, week t adjusted for concurrent co-pollutants and time-weighted lag through t−1. N = 713,918. Single-pollutant plots: DLM_models_krg.png (Figure S6 / Tables S8, S10).

Table 2. G-Formula Estimates Under 20% Weekly Exposure Reduction

Scenario	Prevalence at week 36 (95% CI)	RR (95% CI)	RD, pp (95% CI)	PAF, % (95% CI)
Natural course	6.84 (5.43; 8.90)	1.00	0.00	0.00
PM₂.₅ −20%	6.85 (5.43; 8.93)	1.001 (0.988; 1.010)	0.01 (−0.09; 0.07)	−0.08 (−0.98; 1.21)
NO₂ −20%	6.96 (5.63; 9.36)	1.017 (1.001; 1.030)	0.11 (0.01; 0.21)	−1.67 (−3.00; −0.07)
O₃ −20%	6.58 (5.22; 8.86)	0.961 (0.945; 0.977)	−0.27 (−0.44; −0.14)	3.88 (2.32; 5.60)

Note: Multipollutant parametric g-computation; kriging exposures; 250 bootstrap replicates. Excel exports: 02_Output/G-Form-Multi/Summary_results/*_pct20_point_estimates.xlsx.

Figure 4. Cumulative Preterm Birth Risk (20% Reduction, Multipollutant G-Formula)

Note: Natural course vs 20% proportional reduction applied to all gestational weeks for PM₂.₅, NO₂, and O₃ (panels A–C). Follow-up weeks 28–36.

Figure 5. Critical-Window Heatmap (Single-Week 20% Reduction)

Note: Risk difference by intervention week (columns) and follow-up week (rows). Multipollutant models; point estimates; 20% reduction only.

Supplementary material (reported in Gformula_Santiago_PTB_29092026_Supp.docx)

• Methods S1, Table S1: g-formula estimators and bootstrap.
• Table S2: municipality exposure descriptives (from 7.1).
• Figures S1–S5: IDW distributions, temporal trends, kriging vs IDW, pollutant–covariate correlations.
• Tables S3–S4: weekly IQRs and risk-window descriptives.
• Tables S5–S6, S7–S10, Figure S6: period-average and single-pollutant / IDW DLM results.

Repository-only descriptive figures (not in the main PDF)

These support exploration and the supplement but are not numbered manuscript figures:

• Preterm trends: Preterm_trends_2010_2020.png (7.0)
• Spatial means and 20% density overlays: Map_Exposure_daily_mean_and_density_KRG.png (7.1)
• Regional daily time series: Time_distribution_pm25_no2_o3.png (7.1)

---

📁 Data Availability

Input Data Sources

1. Birth Records: Chilean Ministry of Health (DEIS) vital statistics (2010–2020)

   • Location: 01_Data/Input/Nacimientos/
   • Variables: Gestational age, birth weight, parental characteristics, municipality of residence

2. Air Pollution Data: National air-quality monitoring network

   • Location: 01_Data/Input/Clime_series/
   • Pollutants: PM₂.₅ (beta-attenuation), NO₂ (chemiluminescence), O₃ (UV photometry)
   • Interpolation: Ordinary kriging (primary) and IDW (sensitivity)
   • Spatial unit: Municipal administrative centers (33 urban comunas in the analytic cohort)

3. Temperature Data: CR2MET gridded climate product

   • Processed in: 00_Code/4.0 Climate_data_generate.R
   • Variable: Daily mean ambient temperature (TAD)

4. NDVI: MODIS MOD13Q1 (250 m, 16-day composite)

   • Extraction: 00_Code/3.0 NDVI_EarthEngine_commune_extraction.py
   • Gaps imputed with Kalman smoother

5. Socioeconomic Vulnerability Index (SOVI)

   • Location: 01_Data/Input/SOVI/
   • Categories: Low, medium-low, medium-high

6. Municipal Boundaries

   • Location: 01_Data/Input/district_geo/

Processed Datasets

Main analytical datasets are stored in 01_Data/Output/:

• births_2010_2020.RData: cleaned birth records
• Contamination_Climate_Data_2010_2020.RData: merged pollution and climate series
• births_2010_2020_exposure_weeks.RData: weekly gestational exposure histories
• births_2010_2020_exposure_weeks_lagged.RData: weekly data with DLM lag terms

Descriptive outputs from 7.1 are stored in 02_Output/Descriptives/:

• Table_exposure_commune_PM25_O3_summary.xlsx: municipality min/mean/max (Table S2)
• Table_Annual_Summary_Contaminants_Estimators.xlsx: annual pollutant summaries
• Table_Municipality_Summary_Contaminants_Estimators.xlsx: municipality summaries across estimators
• Map_Exposure_daily_mean_and_density_KRG.png / ..._IDW.png: repository spatial panels (7.1)
• Time_distribution_pm25_no2_o3.png: daily regional means (7.1)
• Histogram_*: Figure 2 and Supplementary Figure S1

Manuscript g-computation (14.x): 02_Output/G-Form-Multi/ (Summary_results/, Figures/, Heatmap/, Bootstrap/).

Extended g-computation (10.x): 02_Output/G-Form/ (17 scenarios, single-pollutant weekly models):

• Summary_results/: point estimates and bootstrap CIs ({stub}_point_estimates.xlsx)
• Interventions/: counterfactual exposure histories (RDS)
• WeeklyEffects/, PopulationEffects/: detailed effect objects
• Bootstrap/{stub}/, Heatmap/{stub}/: replicates and single-week maps

Output stubs follow pollutant and scenario: e.g. pm25_pct10, pm25_pct20, pm25_pct30, pm25_lt20, no2_pct30, o3_pct10.

Note: Individual-level birth records cannot be publicly shared due to Chilean data protection regulations. Aggregated results and analysis code are available in this repository.

---

💻 Reproducibility

System Requirements

• R ≥ 4.0.0
• Python 3 (for NDVI extraction via Google Earth Engine)
• Recommended: ≥ 16 GB RAM; Linux server for parallel g-computation

Required R Packages

Automatically installed via 00_Code/0.2 Packages.R and 00_Code/0.2 Packages_gform.R:

• Data manipulation: tidyverse, data.table, janitor, rio
• Spatial analysis: chilemapas, sf, rnaturalearth, maptiles, tidyterra
• Survival analysis: survival, flexsurv, survminer
• Distributed lag / splines: dlnm, splines, mgcv
• Parallel computing: future, furrr, doParallel
• Visualization: ggplot2, patchwork, ggpubr, ggspatial, RColorBrewer, ragg, scales
• Imputation: imputeTS, zoo

Running the Analysis

1. Setup:

   source("00_Code/0.1 Settings.R")
   source("00_Code/0.2 Packages.R")
   source("00_Code/0.3 Functions.R")

2. Data processing (run in order):

   source("00_Code/1.0 Pollution_process_data.R")
   source("00_Code/2.0 Births_process_data.R")
   source("00_Code/3.1 Temp_NDVI_data.R")
   source("00_Code/4.0 Climate_data_generate.R")
   source("00_Code/5.0 Exposure_data_births.R")
   source("00_Code/6.0 Join_full_data.R")

3. Descriptive analysis:

   source("00_Code/7.0 Descriptive_births.R")
   source("00_Code/7.1 Descriptive_exposition.R")
   source("00_Code/8.0 Correlation_pollulants.R")

4. Distributed-lag models (manuscript Figure 3):

   source("00_Code/9.0 DLM_pollution.R")   # prerequisite lags
   source("00_Code/13.0 DLM_multi_pollution.R")
   source("00_Code/13.1 DLM_multi_plots.R")

5. G-computation (manuscript Table 2, Figures 4–5):

   source("00_Code/14.1 G-Form_multi_build_interventions.R")
   source("00_Code/14.2 G-Form_multi_models.R")   # set GFORM_BOOT_ITER=250 to match the paper
   source("00_Code/14.3 G-Form_multi_plots.R")
   source("00_Code/14.4 G-Form_multi_table.R")

6. Extended g-computation (repository only, optional):

   source("00_Code/10.0 G-Form_build_interventions.R")
   source("00_Code/10.2 G-Form_models.R")
   source("00_Code/10.3 G-Form_plots.R")
   source("00_Code/10.4 G-Form_table.R")

   For server/parallel execution:

   GFORM_EXEC_MODE=server Rscript "00_Code/10.2 G-Form_models.R"

   Run a subset of interventions (by number 1–17):

   GFORM_INTERVENTIONS=12,13,16 Rscript "00_Code/10.2 G-Form_models.R"

Notes on Computation Time

• Birth data processing (2.0): moderate (depends on raw file size)
• Weekly exposure expansion (5.0, 6.0): several hours (large longitudinal dataset)
• DLM Cox models (9.0): ~20–30 minutes per pollutant/method
• G-computation bootstrap (10.2): several hours to days (250–500 bootstrap replicates; parallelized on server)
• Total pipeline: plan for multi-hour to overnight runs on a modern workstation or Linux server

Detailed methodological notes: 02_Output/Notas_G-Formula_resultados.md

---

📖 Codebook

Birth Variables

• id: Unique birth identifier
• com: Municipality code
• name_com: Municipality name
• weeks: Gestational age at delivery (weeks)
• date_nac: Date of birth
• sex: Infant sex (Boy/Girl)
• tbw: Birth weight (grams)
• birth_preterm: Preterm birth indicator (<37 weeks)
• birth_very_preterm: Very preterm (28–31 weeks)
• birth_moderately_preterm: Moderate preterm (32–33 weeks)
• birth_late_preterm: Late preterm (34–36 weeks)

Parental and Context Variables

• age_group_mom, educ_group_mom, job_group_mom: Maternal age, education, employment
• age_group_dad, educ_group_dad, job_group_dad: Paternal age, education, employment
• month_week1, year_week1: Month and year of last menstrual period
• covid: COVID-19 period indicator
• vulnerability: SOVI category (Low, Medium-low, Medium-high)

Exposure Variables

• pm25_krg, no2_krg, o3_krg: Weekly kriging-interpolated concentrations
• pm25_idw, no2_idw, o3_idw: Weekly IDW-interpolated concentrations (sensitivity)
• tad: Weekly mean ambient temperature
• ndvi_full: Municipality-level NDVI (full pregnancy average)
• Lag term (Liw): Time-weighted cumulative lag through prior gestational weeks

---

🔬 Methods Detail

Distributed-Lag Exposure

For gestational week (w \geq 2):

[ L_{iw} = \sum_{s=1}^{w-1} \frac{X_{is}}{w - s} ]

Exclusion Criteria

Births were excluded if:

• Outside urban Metropolitan Santiago (33 comunas: Province of Santiago plus Puente Alto)
• Missing date of birth, gestational age, or municipality
• Maternal age <12 or >50 years
• Gestational age <28 weeks
• Multiple births
• Missing covariates
• Implausible birthweight-for-gestational-age (Alexander et al., 1996)
• Fixed-cohort bias: gestational window not fully observed within 2010–2020

G-Computation Interventions

Main manuscript: proportional reduction with (\pi = 0.20) on all gestational weeks of the target pollutant, multipollutant natural-course models, co-pollutants fixed at observed values:

[ X'{iw} = X{iw} \times (1 - \pi), \quad \pi = 0.20 ]

Single-week reduction (Figure 5): the same 20% rule applied only in week (j); all other weeks remain observed; lag terms recomputed.

Repository extensions (10.x): additional (\pi \in {0.10, 0.30}) and cap rules (X'{iw} = \min(X{iw}, c)) with (c \in {5, 10, 15, 20}) µg/m³ (PM₂.₅) or ppbv (NO₂), fitted under single-pollutant weekly Cox models.

Population metrics at week 36: prevalence, expected cases, RR, RD, AR, and PAF (definitions in supplementary Table S1).

---

🗄️ Repository Structure

Contamination_PTB_G-Formula/
├── 00_Code/                        # Analysis scripts
│   ├── 0.1–0.3                     # Settings, packages, functions
│   ├── 1.0–8.0                     # Data processing and descriptives
│   ├── 9.0–9.1, 11.x, 13.0–13.1, 15.x  # DLM and period Cox models
│   ├── 10.0–10.4                   # Extended g-computation (17 scenarios)
│   ├── 12.x, 14.x, 16.x            # Period and multipollutant g-computation
│   └── old_code/                   # Archived scripts
├── 01_Data/
│   ├── Input/                      # Raw data (not publicly available)
│   └── Output/                     # Processed analytical datasets
├── 02_Output/
│   ├── Descriptives/               # Tables, histograms, maps, time series (see 7.1)
│   │   └── assets/                 # Map figure assets (e.g. warning icon)
│   ├── Models/                     # DLM results and figures
│   ├── G-Form/                     # Extended g-computation (10.x)
│   ├── G-Form-Multi/               # Manuscript g-computation (14.x)
│   ├── G-Form-Period/              # Period-exposure g-computation (12.x)
│   ├── G-Form-Period-Multi/        # Period multipollutant g-computation (16.x)
│   ├── Exposure_PO/                # Period-average Cox (11.x)
│   └── Exposure_PO_Multi/          # Multipollutant period Cox (15.x)
├── 03_Paper/                       # Manuscript and supplementary material
│   ├── Gformula_Santiago_PTB_29092026.docx
│   └── Gformula_Santiago_PTB_29092026_Supp.docx
├── 04_Conference/                  # Conference abstracts
└── README.md

---

⚠️ Important Notes

Data Privacy

Individual-level birth records are confidential and cannot be shared publicly. Researchers interested in data access should contact the Chilean Ministry of Health (DEIS).

Air Quality and Climate Data

• National air-quality network: Chilean Ministry of Environment
• CR2MET: Center for Climate and Resilience Research (CR²)

Citation

If you use this code or methodology, please cite:

Blanco, E., Conejeros, J.D., Bravo, I., Cornejo, F., Osses, A., & Benmarhnia, T. Estimating the perinatal health benefits of hypothetical pollution interventions in Santiago, Chile using parametric g-computation. Under Review. 2026.

---

📧 Contact

For questions about the code or methodology:

• Estela Blanco: estela.blanco@uc.cl
• José Daniel Conejeros: jdconejeros@uc.cl

For data access inquiries:

• Chilean Ministry of Health: https://www.minsal.cl

---

📄 License

This project is licensed under the MIT License. See the LICENSE file for details.

---

🤝 Acknowledgments

This research was supported by FONDECYT de Iniciación en Investigación Nº 11240322 and the Center for Climate and Resilience Research (CR²), FONDAP/ANID 1523A0002. We thank the Chilean Ministry of Health (DEIS) for access to birth records, the national air-quality monitoring network for pollution data, and CR² for climate data.

Data sources:

• Birth records: DEIS, Chilean Ministry of Health
• Air pollution: National air-quality monitoring network (SINCA)
• Temperature: CR2MET v2.5, Center for Climate and Resilience Research, Universidad de Chile
• NDVI: MODIS MOD13Q1 via Google Earth Engine