This chapter orients the injury researcher to important methodologic issues in applying the randomized trial design, and illustrates how the design has been applied in injury research. It provides entry points into the literature on randomized trial methodology. The idea for a randomized trial usually originates from interest in the effects of a particular intervention, which becomes one arm of the trial. Random allocation, the feature that distinguishes a randomized trial from other study designs, is responsible for the substantial advantages of this design. Two issues of randomization are at stake. One is whether the assignment sequence that specifies which subject goes to which intervention arm is truly random. The other is allocation concealment, whether the intervention assignment of a subject is adequately hidden from persons involved in deciding whether and when the subject enters the trial. The data analysis is usually to compare the intervention groups on baseline characteristics.

Administrative and other databases originally created for purposes other than research are increasingly being considered for use in injury research. This chapter provides an overview of the benefits and limitations of use of these electronic databases, categories of uses, types of databases that have been employed in injury research, and some of the methods used. A use of secondary data is the linkage of two or more different databases resulting in a new, enriched, database. Issues that must be dealt with, in addition to those of privacy/confidentiality of information, may include the lack of a unique identifier across all databases, handling of diverse file formats, and the final ownership of the linked product. The level of time and effort required to conduct data linkages depends on the amount and type of information available to conduct a linkage, but always seems to be underestimated in the planning stages of a project.

This chapter concentrates on the potential roles of qualitative research in the conduct of injury research investigations. It presents an overview of methodologic issues in qualitative research with reference to three major qualitative research traditions: ethnography (and ethnographic interviewing), participant observation, and focus groups. Participant observation is commonly used by anthropologists to study a culture in depth. Ethnographic interviews can be nested in participant observation activities but are always conducted in the subject's community setting. Focus groups provide a relatively efficient method to elicit beliefs and attitude from multiple individuals simultaneously. The chapter discusses sampling, data collection, and measures to ensure reliability and validity of qualitative studies. The validity of findings can also be enhanced when the researcher shares the interpretation of data with subjects from the study. Qualitative research methods can be effectively combined with quantitative methods to both derive and test hypotheses regarding human aspects of injury control.

This chapter discusses a number of potentially useful measures with application to cohort studies of injuries. It reviews analytic methods for cohort studies that include crude estimates of risk and relative risk, risk difference or attributable risk, etiologic fraction percent or attributable fraction percent, and multivariate techniques. For illustration purposes, the chapter presents data from a report by Felson and associates on the Framingham Study Cohort that are used to show the effect of impaired vision on hip fracture in older adults. Cohort studies used in injury research are classified into three groups, based on the criteria for assembling the cohort: cohorts based on broad inclusion criteria that do not require an illness or injury (the cohort may represent persons in a defined population), cohorts formed from persons with a non-injury illness, and cohorts which include only persons who have been injured.

During the twentieth century, deaths from infectious diseases have declined dramatically around the world, particularly in industrialized countries. Injury prevention strategies have long existed; however, only relatively recently have these interventions been based on firm scientific evidence and rigorous evaluation. It is the evidence-based approach to advances in injury control that holds the most promise in further reducing the impact of injury on our society. One of the most important milestones in the development of injury research was the publication of Accident Research: Methods and Approaches by Haddon et al. in 1964. Research into violence-related injury has been predominantly conducted by criminologists, psychologists, and sociologists. Injury research must incorporate the scientific advances as it searches for etiologic mechanisms of injury, tests new interventions, especially those based on new technology and laboratory discoveries, and examines the impact of interventions on patient outcomes.