Table 2

Data management principles.*

General
     □ Carefully plan data management well ahead of data collection [2,5-8].
     □ Check for problems early, while it is still possible to correct them [5,8].
     □ Provide staff with appropriate training [3,8,15].
     □ Provide clear lines of authority and responsibility [18].
Data collection instruments
     □ Pre-test all data collection instruments [6,18].
     □ Include the version number and date on each form [17].
     □ Label measurement units on data collection forms [17].
     □ Develop mock tables for results and fill them in with elements from data collection forms to ensure you are collecting all the variables you need [13,14].
     □ Focus efforts on the variables needed for the primary analyses [5,8].
     □ Develop a detailed procedural manual for data collection [8,15,18,24]; keep a log of all decisions that alter procedures.
     □ Use a specific code to indicate data elements that are intentionally blank [4].
Data security, entry and cleaning
     □ Double-enter and verify all data [3-6,9,16].
     □ Develop a data dictionary, including allowable and in-range responses [4-6,15,18].
     □ Store both paper forms and computerized data securely [15,24].
     □ Back up computerized data files regularly [5], keeping offsite copies to safeguard against a system failure [4,15].
     □ Thoroughly check data for missing or potentially erroneous items [1-3,5,8,9,17]. Strategies for data checking include range and consistency checks [1-3,5,8,9,17], checking for missing data [3,5,9,17], between-form consistency checks [2,5,8,9,17], comparing forms to check whether they were collected in the proper sequence, whether forms were entered more than once [16], and whether entered forms matched up with the management database [4].
     □ Never obscure or destroy original data; maintain a clear audit trail of all changes to the data [2,5,8,9,15,17].

*Additional regulations that apply to data integrity and security have been enacted since the time of our study. The Health Insurance Portability and Accountability Act (HIPAA) [25] addresses the security and privacy of health data, and 21CFR Part 11 [26] specifically addresses the reliability of electronic records.

Kruse and Mehr BMC Medical Research Methodology 2008 8:61   doi:10.1186/1471-2288-8-61