This paper will cite two cases to see how the author deals with the "data shortage", turn extremely limited data into high-quality research papers, and dig out unlimited clinical value from limited clinical data. I hope to give some inspiration to make full use of clinical data in the future.

Case 1

Reference: "Cold effect of ambient temperature on scientific and harmonious stroke hospitalization: a large-scale database study in Beijing, China during 20 13 and 20 14 years-using distributed lag nonlinear analysis"

Data source: The daily number of stroke inpatients and daily temperature data in Beijing during 2013.1-20/4.12.31.

Impact coefficient: IF=5.7

The article concludes:

Article decomposition:

Objective: To study the influence of climate change on the number of stroke patients.

Classified data: date and daily visits of stroke patients.

Statistical design: supplementary climate data (daily temperature data obtained by public network)

Model selection: distributed lag nonlinear model, mainly based on the clinical point of 1-extreme cold/extreme heat is easy to induce stroke (statistical language is-the influence of temperature on stroke treatment rate may be nonlinear); Clinical viewpoint 2-external factors may not have immediate or direct influence on the incidence of stroke (statistical language-the influence of temperature on the rate of stroke visit may have lag effect or cumulative effect); Statistical angle-date is the * * * relationship between two variables (case daily outpatient volume and body temperature). Therefore, the model should be selected from the perspectives of clinical specialty and statistical specialty.

Highlights of the article: Looking at the highlights of the article from the perspective of data processing, using extremely limited data dimensions (daily visits to strokes), relying on time and date, and making full use of open free data (temperature), we can find a suitable statistical model (distributed lag nonlinear model) under extremely narrow data dimensions.

Statistical model extension:

The "distributed lag nonlinear" model can be used not only to study the influence of environment/climate on diseases, but also to study the lag effect and cumulative effect of some drugs, treatments and interventions on diseases.

References:

Xia yanluo, Nicholas van halm-Lutterodt, et al. 20 16. Cold Effect of Ambient Temperature on Hospitalization of Ischemic and Hemorrhagic Stroke: A large-scale database study conducted in Beijing, China during 20 13 and 20 14 years adopted a distributed lag nonlinear analysis. Environmental pollution 232 (20 18) 90-96