UCSF

Context

Cutpoints can be used as a threshold for screening symptom(s) that warrant intervention(s) and for monitoring patients' responses to these interventions.

Objectives

In a sample of oncology patients undergoing chemotherapy, study purposes were to determine the optimal cutpoints for low, moderate, and high symptom burden and determine if these cutpoints distinguished among the symptom groups in any demographic, clinical, and stress characteristics, as well as QOL outcomes.

Methods

Total of 1329 patients completed a modified version of the Memorial Symptom Assessment Scale (38 symptoms). Using the methodology of Serlin and colleagues, cutpoints were created using symptom occurrence rates and cancer-specific quality of life (QOL) scores. Cutpoints were validated using measures of stress and resilience and a generic measure of QOL (i.e., Medical Outcomes Study Short Form 12 (SF-12)).

Results

Of the 25 possible cutpoints evaluated, the optimal cutpoint, with the largest between category F statistic, was CP8,15 (Low = 0-8, Moderate = 9-15, High = 16-38 symptoms). Percentage of patients in the Low, Moderate, and High cutpoint groups were 25.3%, 36.3%, and 38.4%, respectively. Significant differences were found among the symptom burden groups in global, cancer-specific, and cumulative life stress (i.e., Low < Moderate < High) and resilience and SF-12 (i.e., Low > Moderate > High) scores.

Conclusion

Our findings provide evidence for clinically meaningful cutpoints that can be used to guide symptom assessment and management. These cutpoints may be used to establish alert thresholds for electronic monitoring of symptoms in oncology patients.