AI Algorithm for Identifying Palliative Care Needs in Ovarian Cancer

Age: 18+

Sex: Any

Trial Phase: Academic

Sponsor: Mayo Clinic

Disqualifiers: Seen by palliative care, Hospice, Under 18

No Placebo Group All trial participants will receive the active study treatment (no placebo)

Compensation: Varies

Number of Visits: Unknown

221 Participants Needed

Trial Summary

Will I have to stop taking my current medications?

The trial information does not specify whether you need to stop taking your current medications.

What data supports the effectiveness of the AI Algorithm for Identifying Palliative Care Needs in Ovarian Cancer treatment?

Research shows that machine learning algorithms using patient-reported outcomes can accurately predict 180-day mortality in women with ovarian cancer, which can help guide end-of-life care decisions. This suggests that AI can effectively identify palliative care needs by recognizing patterns in patient data. ^{1 2 3 4 5}

Is the AI Algorithm for Identifying Palliative Care Needs in Ovarian Cancer safe for humans?

The research articles do not provide specific safety data for the AI Algorithm for Identifying Palliative Care Needs in Ovarian Cancer, but they suggest that using AI to predict mortality in cancer patients is feasible and could help guide care decisions. There is no mention of any safety concerns related to the use of these algorithms in humans. ^{1 3 6 7 8}

How does the AI algorithm treatment for identifying palliative care needs in ovarian cancer differ from other treatments?

This AI algorithm treatment is unique because it uses machine learning to predict the likelihood of mortality within 180 days for women with ovarian cancer, helping to guide palliative care decisions. Unlike traditional treatments, it focuses on improving prognostic accuracy and aligning end-of-life care with patient preferences, rather than directly treating the cancer itself. ^{1 4 9 10 11}

What is the purpose of this trial?

This clinical trial tests an artificial intelligence (AI) algorithm for its ability to identify patients who may benefit from a palliative care consult for gynecologic cancer that has spread from where it first started to nearby tissue, lymph nodes, or distant parts of the body (advanced). A significant delay in referral to palliative care often occurs among patients with cancer. This delay can lead to poorer symptom management, decreased quality of life, and care that does not align with patient goals or values. AI algorithms are computer programs that use step-by-step procedures to solve a problem. In this trial, an AI algorithm is applied to patients' medical records in order to identify patients with a high burden of disease. Information gathered from this study may help researchers learn whether this AI algorithm is useful for identifying patients who could benefit from outpatient palliative care consultation.

Research Team

Rachel D. Havyer, MD

Principal Investigator

Mayo Clinic in Rochester

Eligibility Criteria

This trial is for patients with advanced gynecologic cancer, such as genital warts or ovarian cancer. It's designed to see if an AI algorithm can identify those who would benefit from outpatient palliative care to manage symptoms and improve quality of life.

Inclusion Criteria

Weekly the reviewers will select patients by looking at patients in sorted order starting with the highest score and proceeding down the list and evaluating each patient for exclusion criteria

I have an advanced gynecologic cancer.

Exclusion Criteria

I am under 18 years old.

Patients currently enrolled with hospice

I have not seen a palliative care specialist in the last 75 days.

Timeline

Screening

Participants are screened for eligibility to participate in the trial

2-4 weeks

AI Algorithm Application

Patients' medical records are reviewed for consideration of palliative care consult using AI algorithm

6 months

Weekly reviews

Follow-up

Participants are monitored for outcomes such as palliative care consultations and advanced care planning notes

6 months

Treatment Details

Interventions

• AI Algorithm

Trial Overview The study tests an AI algorithm that reviews electronic health records to pinpoint patients needing palliative care consultations. The goal is to reduce delays in referrals and ensure care aligns with patient goals.

Participant Groups

Group I: Screening (AI algorithm) Experimental Treatment 2 Interventions

Patients' medical records are reviewed for consideration of palliative care consult using AI algorithm QW for 6 months.

Who Is Running the Clinical Trial?

Mayo Clinic

Lead Sponsor

Trials

3,427

Recruited

3,221,000+

Findings from Research

A machine learning model trained on patient-reported outcomes (PRO) data from 245 women with ovarian cancer achieved an accuracy of 79% in predicting 180-day mortality, indicating its potential to improve prognosis estimation.

The model demonstrated good sensitivity (71%) and specificity (80%), suggesting it could effectively guide end-of-life care decisions for ovarian cancer patients, addressing current challenges in treatment delivery.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Predicting 180-day mortality for women with ovarian cancer using machine learning and patient-reported outcome data. Sidey-Gibbons, CJ., Sun, C., Schneider, A., et al.[2023]

Machine learning and network analysis can effectively identify different phases of palliative care based on patient-reported symptoms, using data from 1507 patient records in a New Zealand palliative care service.

Key symptoms like Poor Appetite and Loss of Energy were found to be central in predicting transitions between illness phases, suggesting that these insights could enhance continuous monitoring and digital therapeutic approaches in palliative care.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Intelligent Palliative Care Based on Patient-Reported Outcome Measures. Sandham, MH., Hedgecock, EA., Siegert, RJ., et al.[2022]

In a study involving 12,350 cancer patients, incorporating patient-reported outcomes (PROs) with electronic health record (EHR) data significantly improved the prediction of short-term mortality, with the combined model achieving an area under the curve (AUC) of 0.86 in tertiary oncology and 0.89 in community oncology settings.

The findings suggest that PROs provide valuable prognostic information that EHR data alone cannot capture, enabling earlier and more targeted supportive care for cancer patients.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Development of Machine Learning Algorithms Incorporating Electronic Health Record Data, Patient-Reported Outcomes, or Both to Predict Mortality for Outpatients With Cancer. Parikh, RB., Hasler, JS., Zhang, Y., et al.[2023]

Palliative care (PC) is crucial for improving quality of life and potentially prolonging survival in women with advanced ovarian cancer, yet referrals for PC remain low due to various factors.

A review of 13 articles identified multiple socioecological factors influencing PC referrals, including tumor characteristics, age, marital status, and support systems, indicating that the decision for referral is complex and not solely based on medical conditions.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Multilevel Determinants of Palliative Care Referral in Women With Advanced Ovarian Cancer: A Scoping Review. Cho, S., Goff, BA., Berry, DL.[2023]

Most women with ovarian cancer are diagnosed at an advanced stage, leading to a five-year survival rate of less than 50%, highlighting the need for improved care strategies.

There is a growing emphasis on integrating palliative care early in the treatment of ovarian cancer, with recommendations for targeted education for palliative care clinicians to enhance patient management and decision-making.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Top 10 Tips Palliative Care Clinicians Should Know When Caring for Patients with Ovarian Cancer. Davidson, BA., Moss, HA., Kamal, AH.[2019]

An augmented intelligence tool was developed to predict 30-day mortality risk in cancer patients using socioeconomic and clinical data, demonstrating its feasibility in a large community hematology/oncology practice.

Patients identified as high risk for short-term mortality had a significantly higher event rate of 4.9%, compared to just 0.7% for low-risk patients, indicating the tool's potential to effectively identify those needing urgent interventions or palliative care.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Augmented intelligence to predict 30-day mortality in patients with cancer. Gajra, A., Zettler, ME., Miller, KA., et al.[2021]

A multi-method approach was used to develop a Patient-Reported Outcomes tool tailored for women with endometrial or ovarian cancer undergoing chemotherapy, identifying 21 relevant symptoms from the PRO-CTCAE library.

The final tool includes 44 items that assess the frequency, severity, and impact of these symptoms on daily activities, which can help in monitoring and managing adverse events during treatment.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

A multi-method approach to selecting PRO-CTCAE symptoms for patient-reported outcome in women with endometrial or ovarian cancer undergoing chemotherapy. Christiansen, MG., Pappot, H., Jensen, PT., et al.[2023]

The Ovarian Cancer Survivorship Survey, conducted from 2009 to 2013, aimed to identify factors contributing to long-term survival in ovarian cancer patients, revealing important prediagnostic symptoms that could aid in early detection.

The study highlights a potential link between endometriosis and early-stage ovarian cancer diagnosis, suggesting that understanding these associations may improve future research and patient outcomes.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Design and Implementation of a Comprehensive Web-based Survey for Ovarian Cancer Survivorship with an Analysis of Prediagnosis Symptoms via Text Mining. Sun, J., Bogie, KM., Teagno, J., et al.[2020]

Artificial intelligence tools, like machine learning and natural language processing, can enhance decision-making in palliative care by identifying patients at high risk of mortality and improving prognostic abilities, which can lead to better treatment discussions earlier in the disease process.

While these tools can support essential aspects of palliative care, such as advance care planning, it is crucial to ensure they do not disrupt the vital doctor-patient relationship, especially when communicating sensitive information about prognosis.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

[Artificial intelligence and palliative care: opportunities and limitations.] Peruselli, C., De Panfilis, L., Gobber, G., et al.[2022]

The rate of hospice discharge for metastatic ovarian cancer patients increased from 9.2% in 2004 to 11.1% in 2011, indicating a growing recognition of hospice care as a viable option for end-of-life care.

Documented palliative care services also rose significantly from 2.7% to 10.4% during the same period, while inpatient mortality decreased from 9.6% to 7.4%, suggesting improved care and support for patients with metastatic ovarian cancer.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Trends in hospice discharge, documented inpatient palliative care services and inpatient mortality in ovarian carcinoma. Uppal, S., Rice, LW., Beniwal, A., et al.[2017]

The use of an AI tool to predict 30-day mortality risk in patients with advanced cancer significantly increased palliative care (PC) consults from 17.3 to 29.1 per 1,000 patients per month after its deployment, indicating improved identification of patients needing supportive care.

Hospice referrals also saw a substantial rise, increasing from 0.2 to 1.6 per 1,000 patients per month, demonstrating that the AI tool effectively facilitated timely end-of-life care aligned with patient preferences.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Impact of Augmented Intelligence on Utilization of Palliative Care Services in a Real-World Oncology Setting. Gajra, A., Zettler, ME., Miller, KA., et al.[2023]

References

1. England pubmed.ncbi.nlm.nih.gov

Predicting 180-day mortality for women with ovarian cancer using machine learning and patient-reported outcome data. [2023]

2. United States pubmed.ncbi.nlm.nih.gov

Intelligent Palliative Care Based on Patient-Reported Outcome Measures. [2022]

3. United States pubmed.ncbi.nlm.nih.gov

Development of Machine Learning Algorithms Incorporating Electronic Health Record Data, Patient-Reported Outcomes, or Both to Predict Mortality for Outpatients With Cancer. [2023]

4. United States pubmed.ncbi.nlm.nih.gov

Multilevel Determinants of Palliative Care Referral in Women With Advanced Ovarian Cancer: A Scoping Review. [2023]

5. United States pubmed.ncbi.nlm.nih.gov

Top 10 Tips Palliative Care Clinicians Should Know When Caring for Patients with Ovarian Cancer. [2019]

6. England pubmed.ncbi.nlm.nih.gov

Augmented intelligence to predict 30-day mortality in patients with cancer. [2021]

7. Germany pubmed.ncbi.nlm.nih.gov

A multi-method approach to selecting PRO-CTCAE symptoms for patient-reported outcome in women with endometrial or ovarian cancer undergoing chemotherapy. [2023]

8. United States pubmed.ncbi.nlm.nih.gov

Design and Implementation of a Comprehensive Web-based Survey for Ovarian Cancer Survivorship with an Analysis of Prediagnosis Symptoms via Text Mining. [2020]

9. Italy pubmed.ncbi.nlm.nih.gov

[Artificial intelligence and palliative care: opportunities and limitations.] [2022]

10. United States pubmed.ncbi.nlm.nih.gov

Trends in hospice discharge, documented inpatient palliative care services and inpatient mortality in ovarian carcinoma. [2017]

11. United States pubmed.ncbi.nlm.nih.gov

Impact of Augmented Intelligence on Utilization of Palliative Care Services in a Real-World Oncology Setting. [2023]