CXover.data {rSPARCS}

Generate the Dataset for Case Crossover Analysis

Description

Generate the dataset for case crossover analysis.

Usage

CXover.data(data,date,ID,direction,apart)

Arguments

data

a data.frame containing the date of each case.

date

the name of the variable in the data indicating the date of each case reported to the database.

ID

the name of the variable in the data indicating case ID, if not specified, it will automatically generated starting from 1.

direction

"month4" (default),"pre4" or "after4". With "pre4" (or "after4"), each case day will be matched with same weekdays in previous (or subsequent) 4 weeks. With "month4", each case day will be matched with same weekdays in the same month, which is the most common in literature.

apart

7 (default) or 14. With apart==7, each case day will be 7 days apart from control days in the same month as in the traditional case-crossover design while with apart==14, days will be 14 days apart each other.

Details

Not limited to hospital data, but also applicable to other surveillance data.

Value

dataset

A data.frame ready for the case crossover analysis, with following variables:

ID

same ID represents the same patient.

Date

one case day is matched with 3-4 control days.

status

indicating whether it is a case day or a control day.

References

Zhang W, Lin S, Hopke PK, et al. Triggering of cardiovascular hospital admissions by fine particle concentrations in New York state: Before, during, and after implementation of multiple environmental policies and a recession. Environ. Pollut. 2018;242:1404–1416.

Examples

# similated data
set.seed(2018)
dataset=data.frame(
  patient=1:1000,
  primdiag=sample(390:398,1000,replace=TRUE),
  onset=sample(seq.Date(as.Date("2015/2/1"),as.Date("2016/2/1"),"1 day"),1000,replace=TRUE),
  sex=sample(c("M","F"),1000,replace=TRUE),
  county=sample(c("Albany","New York"),1000,replace=TRUE))

out.data=CXover.data(data=dataset,date="onset",ID="patient")
head(out.data)