Team Clinician Variability in Return-to-Play Decisions
Team Clinician Variability in Return-to-Play Decisions
This survey reports the responses of experienced team clinicians to questions related to nomenclature and clinical decision making in the RTP process. Clinicians chose a variety of responses for each of the injury scenarios. This could be because of a difference in their clinical care or a difference in the interpretation of the word "cleared." However, what may be of more interest is that approximately 35% (n = 22) of respondents would return an athlete to participation in the presence of an elevated risk of acute reinjury or long-term sequelae. In addition, there was limited agreement in the weighting of the factors from the decision-making model. Furthermore, although participants rarely chose "not applicable" for factors in steps 1 and 2 of the 3-step RTP decision-making model, "not applicable" was commonly chosen for factors in step 3 (decision modifiers), with a range of 10% to 45% of participants depending on the particular factor.
The respondents were most consistent in their RTP binary decision (Table 3) when the scenario represented the lowest risk for the athlete (where 95% of respondents would clear an athlete to participate) or the highest risk for the athlete (where only 6% of respondents would clear an athlete to participate). The greatest variability was found in scenarios where the athlete was either at risk of reinjury or long-term sequelae (rows 3 and 4 in Table 3). There are several possible explanations for this finding. The scenarios may not have contained enough specific clinical information to allow the respondents to rank the severity of the injury. In addition, RTP decisions would be different for minor versus major conditions. Finally, the categories of increased risk of acute reinjury or long-term sequelae may be too coarse and should be more granular. This was further investigated in Table 4 where we looked at how respondents changed their answers when given 6 graded choices (Figure 2) rather than the 2 choices "cleared" or "not cleared" (Table 3). Although the overall percentages of respondents that would "clear" an athlete to return to play are similar between Table 3 and Figure 2 in all scenarios, the different percentages observed in Table 4 illustrate that individual respondents' answers often differed depending on the number of choices provided.
We have argued that physicians are best suited to make recommendations based on factors in step 1 and step 2 even though the risk analysis lies outside their formal medical training. However, opinion varies as to the locus of responsibility for step 3 decision modifiers. The current results support the differentiation of step 3 from steps 1 and 2. Within each step of the decision-based RTP model, the factors potential seriousness (of injury) (step 1, 36%, n = 15), type of sport (step 2, 57%, n = 24), and timing and season (step 3, 48%, n = 20) received the highest ranking, whereas some factors in step 3 were selected as "not applicable" (Table 5). Our interpretation of the latter finding is that most clinicians use some nonmedical factors at times, but some clinicians consider 1 or more of the nonmedical factors as either unimportant or not relevant to the decision at other times. This has important implications. At the heart of the RTP decision is the ability to assess the risk of injury to the athlete and the factors in steps 1 and 2 of the RTP decision–based model affecting this risk were generally considered important. Because factors that reflect other athletes' needs/desires (step 3 decision modifiers) were more often considered nonimportant to the RTP decision process, it suggests that sport medicine clinicians may have a more restricted view of the "athlete's best interest" compared with the athlete him/herself. If so, either the generally recommended shared decision-making process may not be applicable in some or all sport medicine contexts or would require a change in culture/legal liability frameworks before it could be effectively implemented.
Beyond step 3, research will eventually need to focus on all 19 factors and the development of criteria for each factor. For example, standardization of functional testing (a factor in step 1) is one area that is receiving more attention. As early as 1995, Mitten and Mitten proposed that sports medicine adapt a similar system to the "industrial rehabilitation" used for workers' compensation cases. Skills needed to complete a worker's tasks are broken down to objective measures of strength, range of motion, and function. Similarly, standardized objective measures for RTP decisions that measure the building blocks of sports movements may eventually provide better risk estimates for reinjury than the current subjective-based methods.
One obvious limitation was that this study used a survey to seek answers to very difficult real-life situations. Different geographical locations, clinical experience with different sports, and different levels of sport are suspected to also influence the responses to this survey. Despite these limitations, these pilot data offer insight to the issues surrounding RTP decisions.
Discussion
This survey reports the responses of experienced team clinicians to questions related to nomenclature and clinical decision making in the RTP process. Clinicians chose a variety of responses for each of the injury scenarios. This could be because of a difference in their clinical care or a difference in the interpretation of the word "cleared." However, what may be of more interest is that approximately 35% (n = 22) of respondents would return an athlete to participation in the presence of an elevated risk of acute reinjury or long-term sequelae. In addition, there was limited agreement in the weighting of the factors from the decision-making model. Furthermore, although participants rarely chose "not applicable" for factors in steps 1 and 2 of the 3-step RTP decision-making model, "not applicable" was commonly chosen for factors in step 3 (decision modifiers), with a range of 10% to 45% of participants depending on the particular factor.
The respondents were most consistent in their RTP binary decision (Table 3) when the scenario represented the lowest risk for the athlete (where 95% of respondents would clear an athlete to participate) or the highest risk for the athlete (where only 6% of respondents would clear an athlete to participate). The greatest variability was found in scenarios where the athlete was either at risk of reinjury or long-term sequelae (rows 3 and 4 in Table 3). There are several possible explanations for this finding. The scenarios may not have contained enough specific clinical information to allow the respondents to rank the severity of the injury. In addition, RTP decisions would be different for minor versus major conditions. Finally, the categories of increased risk of acute reinjury or long-term sequelae may be too coarse and should be more granular. This was further investigated in Table 4 where we looked at how respondents changed their answers when given 6 graded choices (Figure 2) rather than the 2 choices "cleared" or "not cleared" (Table 3). Although the overall percentages of respondents that would "clear" an athlete to return to play are similar between Table 3 and Figure 2 in all scenarios, the different percentages observed in Table 4 illustrate that individual respondents' answers often differed depending on the number of choices provided.
We have argued that physicians are best suited to make recommendations based on factors in step 1 and step 2 even though the risk analysis lies outside their formal medical training. However, opinion varies as to the locus of responsibility for step 3 decision modifiers. The current results support the differentiation of step 3 from steps 1 and 2. Within each step of the decision-based RTP model, the factors potential seriousness (of injury) (step 1, 36%, n = 15), type of sport (step 2, 57%, n = 24), and timing and season (step 3, 48%, n = 20) received the highest ranking, whereas some factors in step 3 were selected as "not applicable" (Table 5). Our interpretation of the latter finding is that most clinicians use some nonmedical factors at times, but some clinicians consider 1 or more of the nonmedical factors as either unimportant or not relevant to the decision at other times. This has important implications. At the heart of the RTP decision is the ability to assess the risk of injury to the athlete and the factors in steps 1 and 2 of the RTP decision–based model affecting this risk were generally considered important. Because factors that reflect other athletes' needs/desires (step 3 decision modifiers) were more often considered nonimportant to the RTP decision process, it suggests that sport medicine clinicians may have a more restricted view of the "athlete's best interest" compared with the athlete him/herself. If so, either the generally recommended shared decision-making process may not be applicable in some or all sport medicine contexts or would require a change in culture/legal liability frameworks before it could be effectively implemented.
Beyond step 3, research will eventually need to focus on all 19 factors and the development of criteria for each factor. For example, standardization of functional testing (a factor in step 1) is one area that is receiving more attention. As early as 1995, Mitten and Mitten proposed that sports medicine adapt a similar system to the "industrial rehabilitation" used for workers' compensation cases. Skills needed to complete a worker's tasks are broken down to objective measures of strength, range of motion, and function. Similarly, standardized objective measures for RTP decisions that measure the building blocks of sports movements may eventually provide better risk estimates for reinjury than the current subjective-based methods.
One obvious limitation was that this study used a survey to seek answers to very difficult real-life situations. Different geographical locations, clinical experience with different sports, and different levels of sport are suspected to also influence the responses to this survey. Despite these limitations, these pilot data offer insight to the issues surrounding RTP decisions.
