Patients with any active drug therapy treatment for cancer receiving care at pilot practice in southeastern Pennsylvania

Patients with any active drug therapy treatment for cancer receiving care at comparison practice in southeastern Pennsylvania

Patients with an evaluation \& management claim attributed to a medical oncology pilot practice in southeastern Pennsylvania

Patients with an evaluation \& management claim attributed to a medical oncology comparison practice in southeastern Pennsylvania

Patients with a new diagnosis of cancer in the past two years

Patient-Centered Oncology Care addresses six domains: track \& coordinate referrals, provide access and communication, identify and coordinate patient populations, plan and manage care, track \& coordinate care, and measure and improve performance.

Pilot of Patient-Centered Oncology Care

Sarah H Scholle, DrPH, MPH

The National Committee for Quality Assurance has worked with the National Coalition for Cancer Survivorship, the American Society of Clinical Oncology, Oncology Management Services, Independence Blue Cross, and RAND, as well as a broader multi-stakeholder advisory group, to define the Patient-Centered Oncology Care model. The purpose of this project was to pilot and evaluate this model. Specific research questions were:

1. Does Patient-Centered Oncology Care improve patient experiences and quality of care? Does it reduce undesirable events like emergency department visits and hospital stays?
2. How does adoption of Patient-Centered Oncology Care vary across a variety of practices and what factors affect adoption?

The demonstration occurred in oncology practices in southeastern Pennsylvania. Practices received implementation support during the 24-month demonstration period. They were evaluated using patient surveys, quality measures, and measures of emergency department and hospital use. Results from these practices were compared in two ways: 1) with their performance before they became oncology medical homes and 2) with other similar practices.

Background. Advances in cancer treatment mean that a growing number of Americans are living with cancer and experiencing it as a chronic, long-term condition. National panels led by consumers have identified the need for improved cancer care in the areas of communication between providers and patients and their families, care planning, attention to nonmedical needs, care coordination and provision of evidence-based treatment. The patient-centered medical home (PCMH) model of care is being widely adopted as a way to provide accessible, proactive, coordinated care and self-care through primary care practices. During active treatment for cancer, the oncology practice is often the primary setting supporting the patient and coordinating cancer treatment. By implementing the patient-centered medical home model, an innovative oncology practice in Pennsylvania has been able to improve access and reduce emergency department visits and hospitalizations for its patients.

Objectives. Building on these recommendations and experience, the National Committee for Quality Assurance has worked with the National Coalition for Cancer Survivorship, the American Society of Clinical Oncology, Oncology Management Services, Independence Blue Cross, and RAND, as well as a broader multi-stakeholder advisory group, to define the Patient-Centered Oncology Care model. We sought PCORI support to pilot and evaluate this model. Specific research questions were:

1. Does Patient-Centered Oncology Care improve patient experiences and quality of care? Does it reduce undesirable events like emergency department visits and hospital stays?
2. How does adoption of Patient-Centered Oncology Care vary across a variety of practices and what factors affect adoption?

Methods. The demonstration occurred in 5 oncology practices in southeastern Pennsylvania. Practices received implementation support during the 24-month demonstration period. They were evaluated using patient surveys, quality measures, and measures of emergency department and hospital use. Results from these practices were compared in two ways: 1) with their performance before they became oncology medical homes and 2) with other similar practices. Patients, clinicians, and health plan leaders helped design the project, and disseminate results. The project used Patient Centered Outcomes Research Institute (PCORI) resources efficiently by building on ongoing efforts. Our evaluation consisted of a retrospective, pre-post study design with a concurrent non-randomized control group of 18 local practices for the utilization and patient experience outcomes. We used difference-in-difference regression models that accounted for practice-level clustering and used functional forms appropriate to the dependent variables. For quality, we compared baseline and follow-up pilot practice performance to national and regional benchmark performance data.

Patient Outcomes. People with cancer are seeking high quality, coordinated and supportive care. The Patient-Centered Oncology Model has the potential to address current gaps in cancer care.

Evaluating the Impact of Patient-Centered Oncology Care

Patient experience survey composite scores were evaluated in the pilot practice group and in the comparison practice group during the intervention period and the follow-up period using a modified version of the cancer CAHPS patient survey.

We calculated survey composite scores on a 0-100 scale using proportional scoring and the summated rating method based on the CAHPS macro.This method calculates the mean responses to each survey item in the composite, after transforming each response to a 0-100 scale (100 representing the most positive response on any given item response scale; 0 representing the least positive). For example, on a Yes/No response scale, if "Yes" represents the most positive response, then Yes= 100 and No = 0; on an Always/Usually/Sometimes/Never response scale, if "Always" represents the most positive response, then Always = 100, Usually = 67, Sometimes = 33 and Never = 0. A higher score means that practices were rated more positively for care on that item.

Quality measure performance rates were evaluated in the pilot practice group during the baseline period and the follow-up period. Data were abstracted from medical records for a sample of patients diagnosed with an invasive malignancy within previous 2 years and with at least 2 visits to the practice in the previous six months at baseline and at 36 months follow-up.

Per member per month hospitalizations, emergency department visits, primary care visits and specialist visits were evaluated in the pilot group and in the comparison group using insurance claims data during the baseline, start-up, intervention and follow-up periods.

American Society of Clinical Oncology

Independence Blue Cross

National Coalition for Cancer Survivorship

Oncology Management Services

RAND

National Committee for Quality Assurance

All-Cause Mortality, Serious, and Other (Not Including Serious) Adverse Events were not monitored/assessed

Patients with any active drug therapy treatment for cancer receiving care at pilot practice in southeastern Pennsylvania

Intervention:

Patient-Centered Oncology Care: Patient-Centered Oncology Care addresses six domains: track \& coordinate referrals, provide access and communication, identify and coordinate patient populations, plan and manage care, track \& coordinate care, and measure and improve performance.

Pilot Practices Patient Survey Cohort

Patients with any active drug therapy treatment for cancer receiving care at comparison practice in southeastern Pennsylvania

Intervention:

No intervention/Usual care

Comparison Practices Patient Survey Cohort

Pilot Practices Utilization Cohort

Patients with any active drug therapy treatment for cancer receiving care at comparison practice in southeastern Pennsylvania

Intervention: No intervention/Usual care

Comparison Practices Utilization Cohort

Patients with a new diagnosis of cancer in the past two years

Intervention:

Patient-Centered Oncology Care: Patient-Centered Oncology Care addresses six domains: track \& coordinate referrals, provide access and communication, identify and coordinate patient populations, plan and manage care, track \& coordinate care, and measure and improve performance.

Pilot Practices Quality Measures Cohort

Total

18-24

25-54

55-64

65-74

75 and older

Unknown

Age, Customized

Female

Male

Sex/Gender, Customized

White

Black/African-American

Asian

American Indian/Alaska Native

Native Hawaiian/Pacific Islander

Other

Hispanic/Latino

Race/Ethnicity, Customized

United States

Region of Enrollment

Sarah Hudson Scholle

To estimate exposure to the intervention on the Access composite score, we used a difference-in-differences model with fixed effects for practices. The dependent variable was the composite score. The intervention effects were represented by the coefficient estimates for survey time period (pre or post) interacted with status (pilot or comparison).

To estimate exposure to the intervention on the Affective Communication composite score, we used a difference-in-differences model with fixed effects for practices. The dependent variable was the composite score. The intervention effects were represented by the coefficient estimates for survey time period (pre or post) interacted with status (pilot or comparison).

To estimate exposure to the intervention on the Shared Decision-Making composite score, we used a difference-in-differences model with fixed effects for practices. The dependent variable was the composite score. The intervention effects were represented by the coefficient estimates for survey time period (pre or post) interacted with status (pilot or comparison).

To estimate exposure to the intervention on the Patient Self-Management composite score, we used a difference-in-differences model with fixed effects for practices. The dependent variable was the composite score. The intervention effects were represented by the coefficient estimates for survey time period (pre or post) interacted with status (pilot or comparison).

To estimate exposure to the intervention on the Exchanging Information composite score, we used a difference-in-differences model with fixed effects for practices. The dependent variable was the composite score. The intervention effects were represented by the coefficient estimates for survey time period (pre or post) interacted with status (pilot or comparison).

To estimate exposure to the intervention on the Overall Rating of Treatment Team composite score, we used a difference-in-differences model with fixed effects for practices. The dependent variable was the composite score. The intervention effects were represented by the coefficient estimates for survey time period (pre or post) interacted with status (pilot or comparison).

Affective Communication Composite-Start-Up

Affective Communication Composite-Follow-Up

Access Composite-Start-Up

Access Composite-Follow-Up

Exchanging Information Composite-Start-Up

Exchanging Information Composite-Follow-Up

Patient Self-Management Composite-Start-Up

Patient Self-Management Composite-Follow-Up

Shared Decision-Making Composite-Start-Up

Shared Decision-Making Composite-Follow-Up

Overall Rating-Start-Up

Overall Rating-Follow-Up

Only the Pilot Practices Patient Survey Cohort and Comparison Practices Patient Survey Cohort were pre-specified to be assessed for this Primary Outcome.

Change in Patient Experience From Baseline to Follow up

Pain assessed, Baseline

Pain assessed, Follow-Up

Pain intensity quantified, Baseline

Pain intensity quantified, Follow-Up

Plan of care for moderate/severe pain, Baseline

Plan of care for moderate/severe pain, Follow-up

Pain addressed defect-free measure, Baseline

Pain addressed defect-free measure, Follow-up

Emotional well-being assessed, Baseline

Emotional well-being assessed, Follow-up

Emotional well-being addressed, Baseline

Emotional well-being addressed, Follow-up

Documented plan for chemotherapy, Baseline

Documented plan for chemotherapy, Follow-up

Chemotherapy intent documented, Baseline

Chemotherapy intent documented, Follow-up

Chemotherapy intent discussion w/ patient, Baselil

Chemotherapy intent discussion with patient, follo

Number of chemotherapy cycles documented, Baseline

Number of chemotherapy cycles documented, Follow-u

Chemotherapy planning defect-free measur, Baseline

Chemotherapy planning defect-free measu, Follow-up

Chemotherapy treatment summary completed, Baseline

Chemotherapy treatment summary completed, Follow-u

Chemotherapy treatment summary to patient, Baselin

Chemotherapy treatment summary to patient, Follow-

Chemotherapy treatment summary to provider, Basel

Chemotherapy treatment summary to provider, Follow

Chemotherapy treatment defect-free measur Baselin

Chemotherapy treatment defect-free measur Follow-u

Tobacco use documented, Baseline

Tobacco use documented, Follow-up

Tobacco use cessation recommended, Baseline

Tobacco use cessation recommended, Follow-up

Tobacco use cessation administered, Baseline

Tobacco use cessation administered, Follow-up

Tobacco defect-free measure, Baseline

Tobacco defect-free measure, Follow-up

Serotonin antagonist prescribed, Baseline

Serotonin antagonist prescribed, Follow-up

Sertonin antagonist/corticosteroids, Ba

Serotonin antagonist/corticosteroids, Follow-up

Aprepitant prescribed, Baseline

Aprepitant prescribed, Follow-up

Anti-emetics defect-free measure, Baseline

Anti-emetics defect-free measure, Follow-up

Infertility risks discussed, Baseline

Infertility risks discussed, Follow-up

Fertility preservation options discussed, Baseline

Fertility preservation options discussed, Follow-u

Patients with a new diagnosis of cancer in the past two years

Patient-Centered Oncology Care: Patient-Centered Oncology Care addresses six domains: track \& coordinate referrals, provide access and communication, identify and coordinate patient populations, plan and manage care, track \& coordinate care, and measure and improve performance.

The pilot practices quality measures cohort were pre-specified to be assessed for this outcome. The pilot practices pulled a sample of patient charts at baseline and follow-up (total of 657 patients). Some patients were eligible for specific measures below and some were not.

Change in Quality of Care From Baseline to Follow up: Percentage of Patients Receiving Recommended Care

To estimate exposure to the intervention on Per Member Per Month Hospitalizations, we used a difference-in-differences model with fixed effects for practices. The dependent variable was Per Member Per Month Hospitalizations. The intervention effects were represented by the coefficient estimates for utilization time period (baseline or follow-up) interacted with status (pilot or comparison).

To estimate exposure to the intervention on Per Member Per Month Emergency Department (ED) Visits, we used a difference-in-differences model with fixed effects for practices. The dependent variable was Per Member Per Month ED Visits. The intervention effects were represented by the coefficient estimates for utilization time period (baseline or follow-up) interacted with status (pilot or comparison).

To estimate exposure to the intervention on Per Member Per Month Primary Care Visits, we used a difference-in-differences model with fixed effects for practices. The dependent variable was Per Member Per Month Primary Care Visits. The intervention effects were represented by the coefficient estimates for utilization time period (baseline or follow-up) interacted with status (pilot or comparison).

To estimate exposure to the intervention on Per Member Per Month Specialist Visits, we used a difference-in-differences model with fixed effects for practices. The dependent variable was Per Member Per Month Specialist Visits. The intervention effects were represented by the coefficient estimates for utilization time period (baseline or follow-up) interacted with status (pilot or comparison).

All-cause hospitalizations, Baseline

All-cause hospitalizations, Start-up

All-cause hospitalizations, Intervention

All-cause hospitalizations, Follow-up

All-cause ED visits, Baseline

All-cause ED visits, Start-up

All-cause ED visits, Intervention

All-cause ED Visits, Follow-up

PCP Office Visits, Baseline

PCP Office Visits, Start-Up

PCP Office Visits, Intervention

PCP Office Visits, Follow-up

Specialist Office Visits, Baseline

Specialist Office Visits, Start-Up

Specialist Office Visits, Intervention

Specialist Office Visits, Follow-Up

Patients with an evaluation \& management claim attributed to a medical oncology pilot practice in southeastern Pennsylvania

Patient-Centered Oncology Care: Patient-Centered Oncology Care addresses six domains: track \& coordinate referrals, provide access and communication, identify and coordinate patient populations, plan and manage care, track \& coordinate care, and measure and improve performance.

Only the Pilot Practices Utilization Cohort and Comparison Practices Utilization Cohort were pre-specified to be assessed for this outcome

Change in Health Care Utilization From Baseline to Follow up

Patients with an evaluation \& management claim attributed to a medical oncology pilot practice in southeastern Pennsylvania

Intervention:

Patient-Centered Oncology Care: Patient-Centered Oncology Care addresses six domains: track \& coordinate referrals, provide access and communication, identify and coordinate patient populations, plan and manage care, track \& coordinate care, and measure and improve performance.

Patients with an evaluation \& management claim attributed to a medical oncology comparison practice in southeastern Pennsylvania

Intervention: No intervention/Usual care

Spots Global Cancer Trial Database for patient care management