EHR-integrated symptom monitoring associates with reduced symptom prevalence and higher mortality risk in cancer patients

This study represents a secondary analysis of a cluster randomized clinical trial conducted within medical oncology clinics at six hospitals located in the Northeast and Southern United States. The population comprised 3,735 patients with gastrointestinal, thoracic, or gynecologic cancer who initiated a new chemotherapy regimen. The intervention involved the implementation of electronic health record-integrated symptom questionnaires, supplemented by patient education, decision support alerts, and clinical reports designed to facilitate symptom management. The comparator arm was not explicitly reported in the available data. The follow-up period for this analysis extended 180 days from the initiation of the chemotherapy regimen.

The primary outcomes assessed included patient-reported symptoms, physical function, overall well-being, and mortality. The analysis revealed a significant temporal improvement in symptom prevalence. The prevalence of severe symptom reports decreased from 5% in week 1 to 1% in week 26, representing a reduction from 1,440 of 30,660 reports in week 1 to 40 of 3,528 reports in week 26. Similarly, the prevalence of moderate symptom reports decreased from 11% in week 1 to 8% in week 26, dropping from 3,522 of 30,660 reports in week 1 to 265 of 3,528 reports in week 26. These reductions indicate a potential benefit of the integrated monitoring system in managing patient symptoms over the course of treatment.

Regarding mortality and physical function, the analysis identified specific symptom severities associated with a higher hazard of death. Moderate physical function deficits were associated with a hazard ratio of 2.07 (95% CI 1.34-3.20; P < .001). Severe physical function deficits were associated with a hazard ratio of 3.39 (95% CI 2.20-5.22; P < .001). Moderate pain was associated with a hazard ratio of 1.43 (95% CI 1.08-1.90; P = .01). Severe pain was associated with a hazard ratio of 1.66 (95% CI 1.21-2.26; P = .001). Moderate dyspnea was associated with a hazard ratio of 1.31 (95% CI 1.02-1.69; P = .04). Severe dyspnea was associated with a hazard ratio of 1.62 (95% CI 1.17-2.24; P = .004). Moderate loss in appetite was associated with a hazard ratio of 1.40 (95% CI 1.02-1.92; P = .04). Severe loss in appetite was associated with a hazard ratio of 2.27 (95% CI 1.55-3.33; P < .001).

Safety and tolerability data were not reported in the available information. There were no specific details provided regarding adverse events, serious adverse events, discontinuations, or general tolerability associated with the intervention or the underlying disease process within the context of this specific analysis. The study design did not allow for a direct comparison of adverse event rates between the intervention and comparator groups as the comparator was not reported.

The results of this analysis should be interpreted with caution regarding causality. As this was a secondary analysis, the observed associations between symptom severity and mortality hazard may reflect underlying disease progression rather than a direct causal link. The findings may help inform programs to monitor and manage symptoms and to deliver targeted interventions that enhance outcomes. However, the study does not establish that the EHR-integrated system itself caused the mortality risk; rather, it highlights the prognostic value of unmanaged severe symptoms in this population.

Key methodological limitations include the lack of reported data on the comparator group and the absence of specific safety reporting. The study was conducted in a specific geographic region (Northeast and Southern US), which may limit generalizability to other settings. The secondary nature of the analysis means that some data points, such as absolute numbers for hazard ratios and specific p-values for the reduction in symptom prevalence, were not fully reported in the input data. These limitations suggest that the results should be viewed as hypothesis-generating rather than definitive proof of efficacy or safety.

Several questions remain unanswered. It is unclear whether the reduction in symptom prevalence was driven by the intervention or by natural disease course and supportive care. The long-term impact of these interventions on overall survival beyond 180 days is unknown. Furthermore, the specific mechanisms by which severe symptoms correlate with higher mortality hazard in this specific cancer population require further investigation. Clinicians should continue to monitor symptoms closely but should not assume that symptom management alone will alter the mortality risk associated with advanced cancer.

In conclusion, this study provides evidence that EHR-integrated symptom monitoring is associated with a reduction in reported symptom prevalence among cancer patients on chemotherapy. However, the association between specific symptom severities and mortality hazard must be interpreted carefully, acknowledging the observational nature of the secondary analysis. The data supports the continued use of systematic symptom assessment but underscores the need for further research to clarify the relationship between symptom burden and survival outcomes.