The identification of a concussion is important in determining if an athlete is to be removed from the game and to reduce the risk of subsequent injury. The decision to remove a player with a suspected concussion is based on a number of factors including the nature and severity of the injury, and the training and/or knowledge of the attending sports medicine professional.

International agreement has seen the strategy for the assessment of a concussion and its severity, moving away from the use of traditional injury severity grading scales1 2 with their emphasis on the loss of consciousness, to a multifaceted approach based on the observation of clinical symptoms, physical signs and cognitive impairment.3 4 In a game or sideline situation the sports medicine professional (athletic trainer, doctor or physiotherapist) is required to make informed decisions about a player’s status under often difficult conditions. Thus, their knowledge and the prioritisation of key information play an important role in the timely identification of a concussion. In New Zealand, where rugby union is a major sporting code, physiotherapists are the primary sports medicine providers for many teams at the community level. However, little is known about what information they, or other sports medicine professionals, use in identifying a concussion.

The Delphi methodology provides a systematic way of determining the collective opinion of a specific cohort and in gaining consensus from the group. This approach has been widely used in medicine,5 6 physiotherapy7 8 and sports medicine9 10 and involves soliciting anonymous input, usually in the form of open question(s) followed by structured question(s), from a defined group of participants. This allows them to evolve their responses over a number or rounds of questioning and provides insight into their knowledge and opinions. It thus provides an ideal methodology to gain a group consensus on important topics. The purpose of this study was to determine the relative importance of signs and symptoms that a selected cohort of sports physiotherapists uses to identify a sports concussion.

METHOD

A two-round Delphi methodology11 12 followed by an educational intervention was used to determine group consensus. The study was conducted with the approval of the University of Otago Human Ethics Committee via email between October 2005 and March 2006.

A cohort of physiotherapists working with sports teams in New Zealand was identified using a purposive sampling strategy to include a spectrum of clinical/teams experience and geographical representation. Round I required participants to identify, in order of importance, up to six signs and symptoms they use to determine if a player has been concussed. A reference group (sports doctor, three sports physiotherapists and a researcher) assigned the responses to categories and formulated the second question (round II), which required the participants to indicate the importance of each category (1 = extremely important, 5 = not important) in their decision making. They were asked to do this with reference to whether they were considering a player in one of two scenarios: a player still in the game (S1) and a player removed from the game (S2). As the reference group doubted whether they would be able to gather any more precise information following the analysis of round II data an educational intervention strategy was introduced whereby the participants were given a copy of a major position paper on sports-related concussion.4 They were asked to read this article (target article) and without further reference to the article rank (1 =  highest, 10 = lowest) in order of importance, a list of 10 categories of information they would use in identifying a concussion. The reference group then compared the participants’ responses to those generated by the research team.

RESULTS

A cohort of 21 (5 females and 16 males) physiotherapists was recruited, with 16 completing the study (76% retention). Participants reported between 1 and 31 years experience as a physiotherapist and averaged 8 (SD 7) years working with sports teams, primarily rugby, and had encountered 4 (SD 3) concussions in the 2005 season. Most (17/21) had recently attended a first aid course in the last 3 years while less than half (9/21) belonged to the national sports medicine association. Participants generated 123 response items (eg, individual responses included “loss of consciousness”, “dizziness”, “memory loss”, “headache” “loss of coordination” and blurred vision”), which the reference group consolidated into eight categories: cognition/orientation; dizziness; headaches; memory; motor dysfunction; nausea/vomiting; state of consciousness; and visual disturbances. The results from round II are presented in table 1 with consensus (80%) being achieved for half of the categories. The data from the post-intervention ratings indicated a lack of concordance between the group and the research team’s expectations (table 2).

DISCUSSION

This cohort of sports physiotherapists identified (by consensus) four categories of items (for both scenarios) as being “very or extremely” important in the identification of a concussion. A major emphasis was placed on the state of consciousness, which most probably includes an obvious loss of consciousness. The group continued to rate this category as the most important following the distribution of the targeted article. It may be argued that the loss of consciousness, although not necessarily an indicator of the severity of the injury,13 is perhaps an obvious sign that a player has been concussed. Following the educational strategy, the group in concordance with the researchers’ rankings, indicated the importance of amnesia/memory problems and dizziness/gait instability/balance difficulties in their decision making, but differed with respect to the importance of headache and the two items relating to vision and limb weakness.

Although not formally assessed, it is assumed that the participants complied with instructions and read and reflected on the target article and its contents. The general observed lack of agreement on certain items cannot be attributed directly to the group’s understanding of the message contained in the article or to the manner in which the concepts are presented, but rather it may very well reflect longstanding beliefs with respect to the identification of a concussion. The need for the strengthening of the link between current research knowledge regarding concussion and clinical practice has been noted by others.14

The present findings suggest the need for further education concerning the key signs and symptoms that are important in the identification of a concussion and/or the revision of the content of current educational programmes. The target article, which has the endorsement of the leading clinicians and researchers working with the management of concussion, may require a more tailored packaging if it is to deliver its message more effectively.