Methods:
The institutional review board approved this study. A clinical
data-warehouse with electronic medical record derived dataset of 94,974
emergent hospital admissions for 57,063 patients within a 6-hospital
network in the Minneapolis/ St Paul area, Minnesota was used. The
encounters spanned a 4-year period ranging from 2012-2016. The hospital
system consists of one 450-bed university tertiary care center and 5 community hospitals ranging from 100-450 beds. Patients were excluded if
they were non-emergent admissions, less than 18
years of age, did not consent to their medical record being used for
research purposes or had less than thirty one days of follow-up mortality data.
We included hospitalizations to all units and services as long as they
met the above criterion. Our database had the complete death record
issued from 2011 onwards for deceased individuals who were born in
Minnesota, had died in Minnesota, or had ever had a permanent address
in the state. In order to avoid bias introduced by repeated measurements from a patient — we randomly sampled the original dataset to include only one
encounter for each patient. SBP were binned into the following group concentrations: <80 mmHg, 80-100 mmHg, 100-120 mmHg, 120-140 mmHg, 140-160 mmHg, 160-180 mmHg, 180-200 mmHg, > 200mmHg. The odds of death or readmission as a function of the final SBP were modeled using logistic regression. The lowest unadjusted rates of adverse events were observed when the systolic blood pressure is in the 120-140 mmHg range and this range was used as the baseline to compute the odds ratio. The primary outcome was death or readmission within 30-days of the index hospitalization. We adjusted the models for age, sex, race and AHRQ comorbidities and estimated glomerular filtration rate (eGFR). Observations with missing values were excluded from the modeling. Analysis was done using R-Studio. R packages tidyverse, icd, nephro, ggplot2, tableone were used.