The use of Patient-Reported Outcomes (PROs) in Cancer Care: A Realistic Strategy

Using natural language processing to analyze unstructured patient-reported outcomes data derived from electronic health records for cancer populations: a systematic review

INTRODUCTION

Patient-reported outcomes (PROs; symptoms, functional status, quality-of-life) expressed in the 'free-text' or 'unstructured' format within clinical notes from electronic health records (EHRs) offer valuable insights beyond biological and clinical data for medical decision-making. However, a comprehensive assessment of utilizing natural language processing (NLP) coupled with machine learning (ML) methods to analyze unstructured PROs and their clinical implementation for individuals affected by cancer remains lacking.

AREAS COVERED

This study aimed to systematically review published studies that used NLP techniques to extract and analyze PROs in clinical narratives from EHRs for cancer populations. We examined the types of NLP (with and without ML) techniques and platforms for data processing, analysis, and clinical applications.

EXPERT OPINION

Utilizing NLP methods offers a valuable approach for processing and analyzing unstructured PROs among cancer patients and survivors. These techniques encompass a broad range of applications, such as extracting or recognizing PROs, categorizing, characterizing, or grouping PROs, predicting or stratifying risk for unfavorable clinical results, and evaluating connections between PROs and adverse clinical outcomes. The employment of NLP techniques is advantageous in converting substantial volumes of unstructured PRO data within EHRs into practical clinical utilities for individuals with cancer.

Synbiotics in Oncology: A Scoping Review Protocol on Their Impact and Outcomes in Cancer Care

Symptom management remains challenging in cancer care. Emerging from nutritional science, nutritional metabolomics has seen exponential growth over recent years, aiming to discern the relationship between dietary habits and health consequences. This protocol aims to present the rationale and methodology for conducting a scoping review to summarize the extent of evidence on synbiotics utilization in cancer symptom management among adults. The scoping review will be undertaken in accordance with the Joanna Briggs Institute (JBI) principles and the research process guided by the PRISMA 2020 scoping reviews extension. The following electronic databases will be searched from the inception: PubMed, Cinahl, Web of Science and Scopus. The authors expect to map the literature regarding the clinical outcomes, including patient-report measures and patient-experience measures, on which the effects of probiotics were tested, and identify potential gaps. This protocol presents a rigorous methodological approach to map the literature on the clinical outcomes that the utilization of synbiotics might improve. This analysis will shape future researchers to examine the efficacy of probiotics on specific clinical outcomes in oncology care. Nurses are uniquely positioned to influence cancer symptom management through the selection and use of appropriate interventions in the field of nutritional supplements, along with nutritional counseling.

Patient-Related Outcome Measures for Oculomotor Symptoms in the Cerebellar Ataxias: Insights from Non-Cerebellar Disorders

In patients with cerebellar ataxia (CA), symptoms related to oculomotor dysfunction significantly affect quality of life (QoL). This study aimed to analyze the literature on patient-related outcome measures (PROMs) assessing QoL impacts of vestibular and cerebellar oculomotor abnormalities in patients with CA to identify the strengths and limitations of existing scales and highlight any areas of unmet need. A systematic review was conducted (Medline, Embase) of English-language original articles reporting on QoL measures in patients with vertigo, dizziness or CA. Pre-specified parameters were retrieved, including diseases studied, scales applied and conclusions drawn. Our search yielded 3671 articles of which 467 studies (n = 111,606 participants) were deemed relevant. The most frequently studied disease entities were (a) non-specific dizziness/gait imbalance (114 studies; 54,581 participants), (b) vestibular schwannomas (66; 15,360), and (c) vestibular disorders not further specified (66; 10,259). The Dizziness Handicap Inventory (DHI) was the most frequently used PROM to assess QoL (n = 91,851), followed by the Penn Acoustic Neuroma Quality-of-Life Scale (n = 12,027) and the Activities-Specific Balance Confidence Scale (n = 2'471). QoL-scores capturing symptoms related to oculomotor abnormalities in CA were rare, focused on visual impairments (e.g., National-Eye-Institute Visual Function Questionnaire, Oscillopsia Functional Impact, oscillopsia severity score) and were unvalidated. The DHI remains the most widely used and versatile scale for evaluating dizziness. A lack of well-established PROMs for assessing the impact of oculomotor-related symptoms on QoL in CA was noted, emphasizing the need for developing and validating a new QoL-score dedicated to the oculomotor domain for individuals with CA.

Care Delivery Interventions for Individuals with Cancer: A Literature Review and Focus on Gastrointestinal Malignancies

BACKGROUND

Gastrointestinal malignancies represent a particularly challenging condition, often requiring a multidisciplinary approach to management in order to meet the unique needs of these individuals and their caregivers.

PURPOSE

In this literature review, we sought to describe care delivery interventions that strive to improve the quality of life and care for patients with a focus on gastrointestinal malignancies.

CONCLUSION

We highlight patient-centered care delivery interventions, including patient-reported outcomes, hospital-at-home interventions, and other models of care for individuals with cancer. By demonstrating the relevance and utility of these different care models for patients with gastrointestinal malignancies, we hope to highlight the importance of developing and testing new interventions to address the unique needs of this population.

What is an "early palliative care" intervention? A scoping review of controlled studies in oncology

INTRODUCTION

Early palliative care (EPC) has been advocated to improve cancer patients' health. However, EPC differs with regard to its elements and target groups. It is not known which parts of EPC contribute to effectiveness for which patient group. This scoping review provides a structured analysis of EPC interventions and outcome measures.

DESIGN

We searched EMBASE, MEDLINE, CINAHL, and CENTRAL up to February 2022. We included randomized controlled trials (RCT), nonrandomized trials, cohort studies (CS), and controlled before-after studies of EPC in adult patients in English, Dutch, and German language. Interventions had to be self-labeled as EPC. Screening and data extraction were performed by two raters. A structured analysis incorporating the TIDieR checklist was performed to describe the elements of the interventions.

RESULTS

We screened 2651 articles, resulting in 40 articles being included: 34 studies were RCT and six studies were CS with a mean sample size of 208 patients. Patients with pancreatic (n = 10) and lung cancer (n = 9) were most often included. Studies reported different reference points for the onset of EPC such as time after diagnosis of incurable cancer (n = 18) or prognosis (n = 9). Thirteen studies provided information about elements of EPC and eight studies about the control intervention. Most frequent elements of EPC were symptom management (n = 28), case management (n = 16), and advance care planning (ACP; n = 15). Most frequently reported outcome measures were health-related quality of life (n = 26), symptom intensity (n = 6), resource use, and the patient's mood (n = 4 each).

CONCLUSION

The elicited heterogeneity of ECP in combination with deficits of reporting are considerable barriers that should be addressed to further develop effective EPC interventions for different groups of cancer patients.

Gaps in Depression Symptom Management for Patients With Head and Neck Cancer

OBJECTIVE

To understand practice patterns and identify care gaps within a large-scale depression screening program for patients with head and neck cancer (HNC).

STUDY DESIGN

Retrospective cohort study.

METHODS

This was a population-based study of adults diagnosed with a HNC between January 2007 and October 2020. Each patient was observed from time of first symptom assessment until end of study date, or death. The exposure of interest was a positive depressive symptom screen on the Edmonton Symptom Assessment System (ESAS). Outcomes of interest included psychiatry/psychology assessment, social work referral, or palliative care assessment. Cause specific hazard models with a time-varying exposure were used to investigate the exposure-outcome relationships.

RESULTS

Of 14,054 patients with HNC, 9016 (64.2%) reported depressive symptoms on at least one ESAS assessment. Within 60 days of first reporting depressive symptoms, 223 (2.7%) received a psychiatry assessment, 646 (7.9%) a social work referral, and 1131 (13.9%) a palliative care assessment. Rates of psychiatry/psychology assessment (HR 3.15 [95% CI 2.67-3.72]), social work referral (HR 1.83 [95% CI 1.64-2.02]), and palliative care assessment (HR 2.34 [95% CI 2.19-2.50]) were higher for those screening positive for depression. Certain patient populations were less likely to receive an assessment including the elderly, rural residents, and those without a prior psychiatric history.

CONCLUSION

A high proportion of head and neck patients report depressive symptoms, though this triggers a referral in a small number of cases. These data highlight areas for improvement in depression screening care pathways.

LEVEL OF EVIDENCE

3 Laryngoscope, 133:2638-2646, 2023.

Development of an addiction recovery patient-reported outcome measure: Response to Addiction Recovery (R2AR)

BACKGROUND

Recovery, a primary goal of addiction treatment, goes beyond abstinence. Incorporating broad domains with key elements that vary across individuals, recovery is a difficult concept to measure. Most addiction-related quality measurement has emphasized process measures, which limits evaluation of treatment quality and long-term outcomes, whereas patient-reported outcomes are richer and nuanced. To address these gaps, this study developed and tested a patient-reported outcome measure for addiction recovery, named Response to Addiction Recovery (R2AR).

METHODS

A multi-stage mixed methods approach followed the Patient-Reported Outcomes Measurement Information System (PROMIS) measure development standard. People with lived experience (PWLE) of addiction, treatment providers, and other experts contributed to item distillation and iterative measure refinement. From an item bank of 356 unique items, 57 items were tested via survey and interviews, followed by focus groups and cognitive interviews.

RESULTS

Face validity was demonstrated throughout. PWLE rated item importance higher and with greater variance than providers, yet both agreed that "There are more important things to me in my life than using substances" was the most important item. The final R2AR instrument has 19 items across 8 recovery domains, spanning early, active, and long-term recovery phases. Respondents assess agreement for each item as (1) a strength, and (2) importance to ongoing recovery.

CONCLUSION

R2AR allows PWLE to define what is important to their recovery. It is designed to support treatment planning as part of clinical workflows and to track recovery progress. Inclusion of PWLE and providers in the development process enhances its face validity. Including PWLE in the development of R2AR and using the tool to guide recovery planning emphasizes the importance of patient-centeredness in designing clinical tools and involving patients in their own care.

Developing and validating utility parameters to establish patient-reported outcome-based perioperative symptom management in patients undergoing prostatectomy: a multicentre, prospective, observational cohort study protocol

INTRODUCTION

Symptom monitoring and alerting based on patient-reported outcomes have proven valuable in a postoperative setting. However, the parameters of the implemented patient-centred symptom management system for patients with prostate cancer (PC) remain unclear. This study aims to develop a perioperative symptom scale (PSS) to monitor symptoms, determine the appropriate timing for symptom assessment and establish intervention criteria for physicians.

METHODS AND ANALYSIS

We will prospectively recruit 387 patients undergoing PC surgery in 3 hospitals. The Chinese version of the MD Anderson Symptom Inventory (MDASI) will be used for longitudinal symptom data collection, presurgery and 1, 2, 3, 5, 7, 14, 21, 28, 42 and 90 days post surgery. A PSS will be generated when symptoms change significantly over time. A linear mixed model will be used to determine appropriate follow-up time points. The functional status determined by MDASI interference can then be used to establish alarm thresholds.

ETHICS AND DISSEMINATION

This study was approved by the Lishui Municipal Central Hospital Ethics Committee on 13 April 2022 (No. LSMCHEC-2022-54) and the Ethics Committee of Huzhou Central Hospital on 5 July 2023 (No. HZCHEC-202306017-01), the Ethics Committee of the First Affiliated Hospital of Huzhou Normal College on 20 June 2023 (No. HZYYEC-2023KYLL055). The latest protocol used in this study was V.2.0, dated on 25 February 2023. Before publication in a peer-reviewed journal, our findings will be presented and discussed at relevant medical conferences.

TRIAL REGISTRATION NUMBER

ChiCTR2200059110.

Patient-reported outcomes in CD30-directed CAR-T cells against relapsed/refractory CD30+ lymphomas

Chimeric antigen receptor (CAR)-T cells targeting CD30 have demonstrated high response rates with durable remissions observed in a subset of patients with relapsed/refractory CD30+ hematologic malignancies, particularly classical Hodgkin lymphoma. This therapy has low rates of toxicity including cytokine release syndrome with no neurotoxicity observed in our phase 2 study. We collected patient-reported outcomes (PROs) on patients treated with CD30 directed CAR-T cells to evaluate the impact of this therapy on their symptom experience. We collected PROs including PROMIS (Patient-Reported Outcomes Measurement Information System) Global Health and Physical Function questionnaires and selected symptom questions from the NCI PRO-CTCAE in patients enrolled on our clinical trial of CD30-directed CAR-T cells at procurement, at time of CAR-T cell infusion, and at various time points post treatment. We compared PROMIS scores and overall symptom burden between pre-procurement, time of infusion, and at 4 weeks post infusion. At least one PRO measurement during the study period was found in 23 out of the 28 enrolled patients. Patient overall symptom burden, global health and mental health, and physical function were at or above baseline levels at 4 weeks post CAR-T cell infusion. In addition, PROMIS scores for patients who participated in the clinical trial were similar to the average healthy population. CD30 CAR-T cell therapy has a favorable toxicity profile with patient physical function and symptom burden recovering to at least their baseline pretreatment health by 1 month post infusion. Trial registration number: NCT02690545.

Interpreting the BREAST-Q for Breast-Conserving Therapy: Minimal Important Differences and Clinical Reference Values

BACKGROUND

The BREAST-Q is an important tool for evaluating patient satisfaction and quality of life in breast-conserving therapy (BCT) patients, but its clinical utility is limited by the lack of guidance on score interpretation. This study determines reference values and the minimal important difference (MID) for the BREAST-Q BCT module.

METHODS

A retrospective review of BCT patients at Memorial Sloan Kettering Cancer Center from January 2011 to December 2021 was performed. Descriptive statistics were used to summarize median BREAST-Q scores. Distribution-based analyses estimated MIDs based on 0.2 standard deviation of baseline BREAST-Q scores and 0.2 standardized response mean of the difference between baseline and 1-year postoperative BREAST-Q scores. MIDs for different clinical groupings based on body mass index, radiation, and reexcision also were estimated.

RESULTS

Overall, 8060 patients were included for determining reference values, and 5673 patients were included for estimating MIDs. Median BREAST-Q scores trended upwards and stabilized by 2 years after surgery for all domains except Physical Well-Being of the Chest, which decreased and stabilized by 2 years. A score interpretation tool, the Real-Time Engagement and Communication Tool, was created based on 25th percentile, median, and 75th percentile scores trajectories. All MID estimates ranged from 3 to 5 points; 4 points was determined to be appropriate for use in clinical practice and research.

CONCLUSIONS

Reference values and MIDs are crucial to BREAST-Q score interpretation, which can lead to improved clinical evaluation and decision making and improved research methodology. Future research should validate this study's findings in different patient cohorts.

The Sarcoma-Specific Instrument to Longitudinally Assess Health-Related Outcomes of the Routine Care Cycle

Patient-based health related quality of life (HRQoL) measurements are associated with an improvement in quality of care and outcomes. For a complex disease such as sarcoma, there is no disease-specific questionnaire available which covers all clinically relevant dimensions. Herein, we report on the development of an electronically implemented, sarcoma-specific instrument to assess health-related outcomes, which encompasses a combination of generic questionnaires tailored to the respective disease and treatment status covering the entire longitudinal care cycle. An interoperable digital platform was designed to provide a node between patients and physicians and to integrate the sarcoma-specific HRQoL instrument with patient and physician-based quality indicators to allow longitudinal structured real-world-time data evidence analytics. This approach enables the prediction modeling of disease, and by attributing cost tags to quality indicators, treatment effectiveness for a given disease will be directly correlated with financial expenses, which may ultimately lead to a more sustainable healthcare system.

Health-related quality of life in breast cancer patients treated with CDK4/6 inhibitors: a systematic review

BACKGROUND

Evaluation of health-related quality of life (HR-QoL) among cancer patients has gained an increasing importance and is now a key determinant of anticancer treatments' value. HR-QoL has been assessed in trials testing cyclin-dependent kinase 4/6 inhibitors (CDK4/6i) in breast cancer (BC), using various questionnaires at different timepoints. HR-QoL reports from BC patients treated with CDK4/6i in the real-world setting are also available.

METHODS

We systematically reviewed the literature, searching for full-length articles, and selected conference abstracts reporting data on HR-QoL in BC patients at any stage and of any molecular subtype treated with abemaciclib, palbociclib or ribociclib.

RESULTS

A total of 533 full-length articles and 143 abstracts were retrieved. After screening for eligibility, 38 records were included (31 clinical trials; 7 real-world reports). Assessment methods were heterogeneous across studies in terms of questionnaires, evaluation timepoints and endpoints. Overall, adding CDK4/6i to endocrine therapy did not worsen patients' HR-QoL, with a positive trend towards pain improvement. Gastrointestinal scores (diarrhea, nausea and appetite loss) statistically favored the control arm among metastatic BC patients receiving abemaciclib, whereas they were superimposable in the early setting. The combination of palbociclib and endocrine therapy showed similar HR-QoL outcomes compared with endocrine therapy alone, but determined better scores compared with chemotherapy. HR-QoL was specifically assessed in premenopausal patients treated with ribociclib, showing similar scores compared with postmenopausal patients.

CONCLUSIONS

Despite methodological heterogeneity does not allow a proper comparison, HR-QoL was generally maintained with CDK4/6i. However, differences between abemaciclib, palbociclib and ribociclib exist and mainly rely on the distinct safety profiles of the compounds. These differences should be acknowledged and taken into account in the clinical practice.

Mental health considerations in patients undergoing complex head and neck reconstruction

PURPOSE OF REVIEW

To summarize recent advances in the psycho-oncology literature, with a focus on findings relevant to the head and neck cancer patient.

RECENT FINDINGS

Patients with cancer are at an increased risk of suicide and self-harm. Head and neck cancer patients are among the highest risk compared with other cancer patients. Unique challenges pertaining to disfigurement and voicelessness may, in part, explain these observations. Patient-reported outcome measures can be used to help identify high-risk patients.

SUMMARY

Psychosocial support needs are highest for head and neck cancer patients. Patient-reported outcome measures should be integrated within clinical workflows to identify high-risk patients.