Intermediate Epidemiology by Manya Magnus

Author:Manya Magnus
Language: eng
Format: epub
Publisher: Jones & Bartlett Publishers, Inc.
Published: 2014-04-15T04:00:00+00:00

* All calculations assume a type I error rate of 0.05. The effect of interest is specified as a risk ratio. Study size is reported per treatment arm, and a 20% dropout rate is assumed for calculating the needed recruitment.

Reproduced from Velentgas P, Dreyer NA, Nourjah P, Smith SR, Torchia MM, eds. Developing a Protocol for Observational Comparative Effectiveness Research: A User’s Guide. AHRQ Publication No. 12(13)-EHC099. Rockville, MD: Agency for Healthcare Research and Quality; January 2013. www.effectivehealthcare.ahrq.gov/Methods-OCER.cfm.

One of the easiest ways to get to know your data is to develop table shells (sometimes called dummy tables), into which you will transfer your data. These tables are often in the same basic structure as the structure you will use to report your data for a final report or manuscript but will have more variables listed than usually are published. Others prefer to use printouts or e-files to run univariate analysis (PROC FREQ or PROC UNIVARIATE in SAS, for example, or tab1 or tabstats in Stata) on all the available or relevant variables. This is useful for some, but paring it down to table shells helps keep the data organized and facilitates review of the statistical output. One way to start is to make a table shell that will systematically collect summaries of all of the key variables. For example, Table 7-22a and Table 7-22b show table shells that might guide a univariate analysis of a study on access to flu shots, for which the data were provided by telephone survey:

Table 7-22a Sample Univariate Table Shell (Categorical Variables)

Variable* n %

Gender

Male

Female

Age (years)

18–25

26–35

36–45

46–55

Annual household gross income (USD)

<$10,000

$10,000–19,999

$20,000–29,999

$30,000–39,999

≥$40,000

Has health insurance

No

Yes

Location of residence

Urban

Suburban

Rural

Clinical information

Contraindication to flu shot (e.g., egg allergy)

Previous flu shot experience

Ever

At least one within past 6 months

At least one within past year

At least one within past 2 years

Characteristics associated with flu shot recommendation

Asthma

Diabetes

Chronic lung disease

Pregnant women

Age older than 65

People who live with or care for others who are at high risk of developing serious complications

* Note: that the categorization for each of the continuous variables will be determined by the distribution of the data derived from the sample. This is a small selection of variables only; a true analysis would have many more variables described. It can also be useful to add a row under each variable’s levels indicating the proportion of missing data, as this will aid you later on in deciding what analytic approaches are needed and what limitations must be considered when you are writing the manuscript.

Table 7-22b Sample Univariate Table Shell (Continuous Variables)

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.