8,700 Studies Reviewed. 87.0% Found Biological Effects. The Evidence is Clear.
Shield Your Body
Whole Body / General

(CE, DE, IU, ME, MO, PN, VO)

Bioeffects Seen

Authors not listed · 2024

Share:

Large-scale health studies reveal patterns invisible in smaller research - a lesson EMF science desperately needs.

Plain English Summary

Summary written for general audiences

This study developed a risk prediction tool to identify patients most likely to develop lung complications after surgery, using data from over 86,000 patients across 114 countries. The model accurately predicted which patients would experience pneumonia or breathing problems within 30 days of their operation. This tool could help hospitals better prepare resources and inform patients about their individual surgical risks.

Why This Matters

While this surgical risk prediction study doesn't directly address EMF exposure, it highlights an important principle we see repeatedly in EMF research - the power of large-scale data collection to reveal health patterns that smaller studies might miss. The researchers analyzed over 86,000 patients across 114 countries to develop their predictive model, demonstrating the kind of comprehensive approach we need more of in EMF health research. Unfortunately, most EMF studies involve far smaller sample sizes, making it harder to detect subtle but significant health effects. The reality is that EMF exposure, like surgical complications, may affect only a small percentage of the population - but when you're talking about billions of people using wireless devices daily, even a 2-4% complication rate translates to millions of affected individuals. This study's methodology shows what's possible when researchers commit to large-scale, international collaboration rather than relying on industry-funded studies with limited scope.

Exposure Information

Specific exposure levels were not quantified in this study.

Cite This Study
Unknown (2024). (CE, DE, IU, ME, MO, PN, VO).
Show BibTeX
@article{ce_de_iu_me_mo_pn_vo_ce3730,
  author = {Unknown},
  title = {(CE, DE, IU, ME, MO, PN, VO)},
  year = {2024},
  doi = {10.1016/S2589-7500(24)00065-7},
  
}
DOI: 10.1016/S2589-7500(24)00065-7PubMed: 38906616

Quick Questions About This Study

The model showed good accuracy with area under the curve scores of 0.773-0.786 in development and 0.716-0.746 in validation studies. This means it correctly identified high-risk patients about 75% of the time, which is considered acceptable for clinical use.
Pulmonary complications occurred in 2.0% of patients in the original dataset, 3.9% during COVID-19 pandemic surgeries, and 4.7% in pre-pandemic UK/Australian data. These rates varied based on timing and geographic location of the surgeries.
Both XGBoost and LASSO regression performed similarly with area under the curve scores around 0.785-0.786. However, LASSO was chosen for the final model because it was more explainable to doctors and required fewer variables.
The study included 1,158 hospitals across 114 countries for model development, plus 726 hospitals in 75 countries and 150 hospitals in 3 countries for validation. This massive international scope strengthened the model's reliability.
The final GSU-Pulmonary Score uses ten simple predictor variables that doctors can easily assess before surgery. These are basic patient characteristics and medical factors available during routine pre-surgical evaluation, making the tool practical for widespread use.