This is an open-access article distributed under the terms of the Creative Commons Attribution-Noncommercial-Share Alike 3.0 Unported, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Case report form (CRF) is a specialized document in clinical research. It should be study protocol driven, robust in content and have material to collect the study specific data. Though paper CRFs are still used largely, use of electronic CRFs (eCRFS) are gaining popularity due to the advantages they offer such as improved data quality, online discrepancy management and faster database lock etc. Main objectives behind CRF development are preserving and maintaining quality and integrity of data. CRF design should be standardized to address the needs of all users such as investigator, site coordinator, study monitor, data entry personnel, medical coder and statistician. Data should be organized in a format that facilitates and simplifies data analysis. Collection of large amount of data will result in wasted resources in collecting and processing it and in many circumstances, will not be utilized for analysis. Apart from that, standard guidelines should be followed while designing the CRF. CRF completion manual should be provided to the site personnel to promote accurate data entry by them. These measures will result in reduced query generations and improved data integrity. It is recommended to establish and maintain a library of templates of standard CRF modules as they are time saving and cost-effective. This article is an attempt to describe the methods of CRF designing in clinical research and discusses the challenges encountered in this process.
Keywords: Case report form, completion guidelines, case report form design, electronic case report form, standard templates
A case report form (CRF) is designed to collect the patient data in a clinical trial; its development represents a significant part of the clinical trial and can affect study success.[1] Site personnel capture the subject's data on the CRF, which is collected during their participation in a clinical trial. The International Conference on Harmonization Guidelines for Good Clinical Practice define the CRF as: A printed, optical or electronic document designed to record all of the protocol – required information to be reported to the sponsor on each trial subject.[2]
Case report form designing requires enormous planning and attention to minute detail. Designing a CRF is crucial in a clinical trial as it will aid in assessing the safety and efficacy of the medicinal product accurately. CRF should be designed for optimal collection of data in accordance with the study protocol compliance, regulatory requirements and shall enable the researcher test the hypothesis or answer the trial related questions.
A well-designed CRF should represent the essential contents of the study protocol and in an ideal situation, CRF is designed once the study protocol is finalized. It can be prepared either concurrently along with the protocol development, but may result in many versions, and hence needs to be version controlled. Timing of the design process will also play an important role as both the approaches have pros and cons.
It is increasingly recognized that the design of the CRF (paper form/electronic form) is a key quality step in ensuring the data required by the protocol, regulatory compliance and/or safety needs/comments, study-specific hypothesis attributes, site work flow, and cross-checking of data items within a form or across different forms are addressed.[3] The CRF used in clinical research reduces messy clinical realities to round integers and categorical answers.[4]
This article is an attempt to describe the methods of CRF designing in clinical research, discusses the challenges encountered and measures to be taken to prevent the occurrence of issues in its development.
There are two types of CRFs used in clinical research, that is, traditional paper CRF and improvised electronic CRF (eCRF). Paper CRF is the traditional way of data capture and a better option if studies are small or vary in design, whereas eCRFs are considered if studies are large with similar designs.[5]
In the current global scenario, eCRFs are preferred over paper CRFs as they are less time-consuming, and also encourage the sponsor/pharmaceutical company to carry out large multicentric studies at the same time due to the ease of administration. It is designed in such a way that data entry can be done with zero/minimal errors. Moreover, the regulatory authorities are readily accepting submissions in which validated electronic data capture (EDC) systems are used.[6]
While designing an eCRF, repetitive data such as protocol ID, site code, subject ID, and patient initials will be generated by the system automatically from the first page to all others, thus ensuring no duplication of CRF pages.
In eCRF, linking the data between two related pages of CRFs becomes easy and quick. They have built-in edit checks tagged to each data field as well as to the CRF as a whole. Therefore, majority of data cleaning activities will take place during the completion of the eCRFs, thus reducing the time and effort required by data management personnel. Instant query resolution reduces the time spent on obtaining the clarification from the site/investigator and hence, clean data is obtained much quickly, resulting in timely database lock, faster regulatory submission, and subsequent approval.
Designing a paper CRF is a tedious job that could result in data errors and wrong conclusions, requiring meticulous attention to minimize duplication of CRF pages. Chances of error during data transfer from the source document to paper CRF are common. Moreover, for studies with large sample size if traditional method of data collection through paper CRFs is opted, then manual data cleaning may be a major concern. However, this method may not require user training and system validation as in the case of EDC systems, where such things are essential before implementing it. Despite their many advantages, eCRFs have not been accepted widely. Main reasons behind this are lack of available on-site technology, investigators’ lack of motivation, complexity of installation, and maintenance of the software and high investment cost.
Designing a CRF is an art that should to be based on scientific practices and the design should be implemented keeping the end-user (the one who enters data in the CRF) in mind. While designing, all important sections of the CRF should be included with care; always it is worth to remember that insufficient/inaccurate data collection would prove expensive during analysis. Hence, it is advisable to have a standard operating procedure for CRF preparation and to follow best practices of CRF designing.
Primary objective of CRF designing is to gather complete and accurate data by avoiding duplication and facilitating transcription of data from source documents onto the CRF. CRF should be designed with the primary safety and efficacy endpoints as the main goal of data collection.[6] Ideally, it should be well-structured, easy to complete without much assistance and should collect data of the highest quality. Always minimum amount of data needed to answer the study hypotheses should be collected avoiding collection of elaborate, unimportant information. For ordinal data, to ensure uniformity and clarity among raters, adequate explanation should be provided adjacent to the CRF fields. Capturing the same piece of data in more than one place (duplication) on the CRF should also be avoided. In other words, CRF should collect data in sufficient detail without ambiguity and at the same time, should avoid redundancy and avoid capture of unwanted details. Hence, striking the perfect chords to ensure balance between effective data collection and structuring the CRF to support accurate data entry is essential. Collecting the data in the coded form whenever possible is ideal as it facilitates data entry (at CRF and at the database levels) and helps the statistician in data interpretation and analysis.
Important part of the CRF is an informative header and footer, which can be customized.[7] In general, the header includes protocol ID, site code, subject ID, and patient initials. Whereas, the footer includes investigator's signature, date of signature, version number, and page number.
In order to enhance easy reading/understanding and accurate data entry, an uncrowded CRF layout should be preferred. Placing too many details on the same page, makes the CRF look cluttered and makes data entry difficult, which eventually leads to increase in data discrepancies.
Case report form design should be standardized to address the needs of all those who handle the data such as investigator, data manager, biostatistician, clinical research monitor/coordinator, database developer/programmer and data entry personnel etc. An effective CRF design would always be user friendly. Moreover, it should capture legible, consistent and valid data, thereby, reducing query generations.[7] While designing the CRFs, design standards should be adhered to for improving the quality of data collected. Hence, data should be organized in a format that facilitates data analysis and makes it simplified.
The following points are to be borne in mind while designing a CRF:
Use of consistent formats, font style and font sizes throughout the CRF booklet Selection of portrait versus landscape versus combination layouts Use of clear and concise questions, prompts, and instructionsVisual cues, such as boxes that clearly indicate place and format of data to be recorded should be provided to the person recording the data as much as possible
Using the option of “circling of answers” should be limited as it's hard to interpret; instead check boxes would be appropriate
Clear guidance about skip patterns like what to skip and what not to skip should be mentioned at appropriate places
Skips (are instructions provided in the CRF page to maintain the connectivity between pages) should be kept to a minimum by the placement of questions to avoid confusions
Provide boxes or separate lines to hold the answers. This indirectly informs the data recorder where to write/enter the response and helps to differentiate it visually from the entry fields for other questions
Separate the columns with thick lines Provide bold and italicized instructions Minimize free text responses Position only specified density of questions on each page Page numbering if necessary, should be consistent throughout Avoid using “check all that apply” as it forces assumptions about the clinical data Specify the unit of measurement Indicate the number of decimal places to be recorded Use standard data format (e.g., dd/mm/yyyy) throughout the CRFUse precoded answer sets such as yes/no, male/female, method of administration of medicine, and severity of adverse event (AE) (mild/moderate/severe) wherever possible
Not to split modules/sections (a set of one or more related groups of questions that pertain to a single clinical study visit) like, for example, AE section should not be split and laid across pages such that information related to a single AE will have to be collected from different pages
Use “no carbon required (NCR)” copies to ensure exact replica of CRFUse instructions including page numbers where data has to be entered (e.g., during a follow-up visit, the investigator is supposed to record whether any AE has occurred and if occurred, details of the AE has to be recorded in the AE module. Hence, the field corresponding to this question on the module for the particular visit would be having the options “yes” or “no”. There should be an instruction “If ‘yes’, please provide the information in the AEs page (page no. XX)” as shown in Figure 1 .
A sample case report form (CRF) page. An adverse event page of CRF is depicted showing codes, and skips questions
Table 1 provides the comparison between well-designed and poorly designed CRF. In case of poorly designed CRF, by placing a single line for required response results in variations in the investigator's responses from site to site. On the other hand, separate lines and boxes are provided in the well-designed form, which gives the visual cues about what is expected as a response and thereby, reduces the unnecessary queries. Usually, boxes are used for entering dates and the date format (i.e., dd/mm/yyyy) should be consistent on all pages of CRF. Units and decimal points should be displayed for vital sign records, which clarify the user about the expected values and also facilitates the data interpretation and reduces manipulation during analysis procedures. Figures Figures2 2 and and3 3 are examples of poorly designed CRFs. Poor CRF design results in frequent database modification thus affecting the study timelines. Data need to be collected in a way that does not introduce bias or errors. Collection of a large amount of data will result in wasted resources in collecting and processing it. Questions in the CRF should be clear and unambiguous to avoid unnecessary confusions.[8]
Illustrating a well-designed and poorly designed data fields imparting the significance of visual cues to help the site personnel to understand the format
