Analyzing Consistency and Injury Risk Sensitivity in Cricket Athletes: Session-RPE for Workload Monitoring

Objective: The present study aimed to verify two key hypotheses: (1) whether the session rating of perceived exertion (sRPE) can serve as a reliable indicator for workload monitoring in cricket and (2) to compare the injury risk sensitivity of sRPE-derived indicators—including the coupled and uncoupled acute:chronic workload ratio (ACWR), exponentially weighted moving average (EWMA), and robust exponential decreasing index (REDI)—across three pre-specified latency periods (no latency, 7-day latency, and 14-day latency), and to identify the optimal indicator and latency period for cricket injury risk assessment. Material: Twenty-four elite female cricket athletes from the Chinese National Women’s Cricket Team were monitored during daily training throughout the Los Angeles Olympic Games preparation period. Methods: Correlation analysis, Kendall’s tau correlation coefficients, and Bland–Altman plots were employed to assess the relationships and consistency between sRPE and various workload indicators. ROC curves were constructed to compare the performance of sRPE-derived indicators for injury risk across the three pre-specified latency conditions. Results: sRPE and its derived indicators exhibited significant correlations with health status indicators (sleep, fatigue, delayed onset muscle soreness (DOMS), stress, mood, and resting heart rate (RHR)) and physiological and biochemical indicators (testosterone, cortisol), with the majority of these correlations reaching the 0.01 significance level (p < 0.01). Kendall’s tau correlation coefficients and Bland–Altman plots revealed that sRPE and ACWR based on EWMA (hereafter referred to as EWMA) had moderate correlations with health status indicators, while ACWR based on REDI (hereafter referred to as REDI) showed a strong correlation with such indicators. sRPE and its derived indicators were strongly correlated with physiological and biochemical indicators (Kendall’s tau > 0.8) with good consistency, as the majority of scattered points fell within the limits of agreement (mean difference ± 1.96 × standard deviation, MD ± 1.96 × SD). Analysis of injury risk sensitivity indicated that the 7-day latency model yielded the highest average area under the curve (AUC = 0.85). Among all indicators, REDI and EWMA achieved the highest AUC values (AUC = 0.665 and 0.667, respectively; p < 0.001). Notably, EWMA exhibited optimal performance in the 7-day latency time series (AUC = 0.859, >0.80), followed by REDI under the 7-day latency condition (AUC = 0.857). Conclusion: EWMA with 7-day latency is a more sensitive indicator for detecting injury risk.