Memorial Sloan Kettering Cancer Center June 09, 2017 Ms. Seem Verma 1\dbnlinistrator Centers for Medicare & Medicaid Services Department of Health and Human Services PO Box 8016 Baltimore, MD 21244-80 16 Re: CMS-1677-P - PCHQR Program, Operating Room (O.R.) and Non-O.R. Issues Dear 1\dministrator Verma: Memorial Hospital for Cancer and 1\llied Diseases (Memorial) appreciates the opportunity to comment on the Center for Medicare & Medicaid Services' (CMS) Inpatient Prospective Payment System (IPPS) Proposed Rule for FY 2018. 1\s the oldest and largest hospital in the world devoted solely to the research and treatment of cancer, Memorial is committed to exceptional patient care, innovative research, and outstanding educational programs. Given our unique focus on cancer care, our comments on the proposed rule are centered on the PPS-Exempt Cancer Hospital Quality Reporting Program and proposed changes to the status of certain procedure codes that could have an enormously unfortunate impact on whether procedures are correctly grouped as autologous and allogeneic transplant. PPS-Exempt Cancer Hospital Quality Reporting (PCHQR) Program Memorial is one of 11 PPS-Exempt Cancer Hospitals (PCHs) nationally that are subject to the PCHQR Program. We appreciate the opportunity to provide feedback and are pleased to offer comments intended to ensure that the clinical quality measures in the program are valid, reliable and useful tools for assessing the quality of cancer care. Removal ofTopped Out Measures Given the evaluation demonstrating that the PCH-01/NQF #0223, PCH-02/NQF#0559, and PCH-03/NQF #0220 measures meet the established criteria for "topped out" measures, Memorial supports the removal of these measures from the PCHQR Program. Retiring these measures Jorge Lopez, Jr. Executive Vice President & General Counsel Office of General Counsel 1275 York Ave, MG16A, New York, NY 10065 T (212) 639-5980 0 F (212) 717-3079 NCI -designated Comprehensive Cancer Center Ms. Seem Venna Centers for Medicare & Medicaid Services Department of Health and Human Services Page 2 of6 from the program will allow PCH resources to be focused on reporting newly adopted measures, including outcome measures, which better align with program goals. Proposed New Quality Measures Beginning with the FY 2020 Program Year At Memorial, patients' quality of life and honoring patient preferences about end-of-life care are of paramount importance, and we greatly appreciate CMS' focus on quality measures designed to assess end-of-life care. While Memorial is conceptually supportive ofthe addition of the four proposed end-of-life measures (NQF #0210, NQF #0213 , NQF #0215, and NQF #0216), the specifications for all of these measures require updating, making it difficult to offer unqualified support for their adoption. Further, a careful review of these measures has identified additional, important considerations that we believe should be addressed prior to finalizing measures #021 0, #0213 , and #0215 for adoption. Memorial recommends that these quality measures be updated to allow for review of current claims-based measure specifications with up-to-date code lists (e.g., using International Classification of Diseases, Tenth Revision, Clinical Modification (ICD10-CM) and Current Procedural Terminology (CPT), as well as to address the additional considerations identified below: • NQF #021 0 - Proportion of Patients Who Died from Cancer Receiving Chemotherapy in the Last 14 Days of Life Memorial is committed to relieving suffering and improving quality of life for patients throughout the course of their care, taking into account patients' unique circumstances, values, and preferences. Measuring the proportion of patients who died from cancer and received chemotherapy in the last 14 days of life provides an important indicator of quality of care, and requiring reporting of this measure may contribute to reducing unnecessary treatment and prompt improvements in patients' quality of life. However, the measure should be reported in a manner that recognizes that a performance rate of zero is not the goal and may, in fact, be undesirable; palliative chemotherapy is appropriate for some patients and other patients may wish to continue to pursue curative treatment until the end-of-life, including by participating in clinical trials. Accordingly, we encourage the measure developer to consider modification ofthe measure to ensure that it does not discourage chemotherapy delivered for palliation (e.g., for painful bone metastases) and recognizes the importance of patient preferences, including participation in clinical trial protocols. We encourage CMS and the measure developer to work with the Alliance of Dedicated Cancer Centers (ADCC), of which Memorial is a member, to work through potential solutions to this issue. • NQF #0213 - Proportion of Patients Who Died from Cancer Admitted to the ICU in the Last 30 Days of Life Memorial supports an assessment of the proportion of intensive care unit (ICU) admissions during the last 30 days of life. Again, however, there are clearly circumstances under which ICU admission may be appropriate, e.g., when treating bone marrow transplant patients with curative intent, and an extremely low a rate of ICU admission could, in certain circumstances, be Ms. Seem Verma Centers for Medicare & Medicaid Services Department of Health and Human Services Page 3 of6 indicative of low quality care. Consequently, without exclusions from the denominator or riskadjustments to account for patient characteristics, this measure is not suitable for public reporting or for introduction into programs where measure performance is tied to reimbursement. While we recognize that these factors are not necessarily prohibitive of its initial introduction into the PCHQR program, the measure developers may want to consider ways to address this issue going forward. As noted by the ADCC, Memorial also encourages CMS to provide hospitals with confidential performance data that stratifies rates between ICU admission of patients to the attributed hospital, as compared to the rate of admission of patients to an ICU at an outside provider. This information may be useful in designing quality improvement efforts. • NQF #0215- Proportion of Patients Who Died from Cancer Not Admitted to Hospice Assessing the proportion of patients who are admitted to hospice is a useful metric of whether patients are receiving adequate referrals to services for pain management, symptom control, and other services designed to promote good quality of life for patients with terminal illness. However, Memorial urges the measure developers to expand this measure to include hospitalbased palliative care services, which are particularly important for patients seeking access to lifeprolonging treatment and clinical trials, in the numerator. The Palliative Medicine Service at Memorial is robust, and we incorporate the principles of palliative care into the treatment of all patients with cancer, often from the time of diagnosis. This approach allows members of the Palliative Medicine Service to work alongside treating oncologists, seeking to ease distress while remaining mindful of unique patient and family characteristics, values, beliefs, and culture. To provide the best palliative care possible, the service may call on individuals from specialties such as interventional/anesthesia pain management, rehabilitation medicine, psychiatry and behavioral science, integrative medicine, chaplaincy, social work, pharmacy, wound care, radiation oncology, and neurology. This interdisciplinary approach improves quality of life for patients as well as their families throughout the course of care, and we urge CMS and the measure developer to modify the measure to include hospital-based palliative care services into the numerator before finalizing its adoption for the PCHQR program. Without this modification, Memorial and other hospitals offering significant palliative care services could be inaccurately perceived as failing to offer patients adequate end-of-life services. • NQF #0216 - Proportion of Patients Who Died from Cancer Admitted to Hospice for Less Than Three Days Memorial supports the inclusion of this measure in the PCHQR program. For the reasons noted above, modifying the measure to incorporate the receipt of palliative care services for less than three days would contribute to a fuller understanding of whether patients who died from cancer were offered adequate services for pain management, symptom control, and other supportive services early enough in their course of care. Ms. Seem Verma Centers for Medicare & Medicaid Services Department of Health and Human Services Page 4 of6 In summary, we commend CMS for its focus on measuring, and thereby potentially spurring improvements in, end-of-life care. However, particularly forNQF Measures #0210, #0213, and #0215 , Memorial believes that refinements to the measure design are necessary before finalizing their inclusion in the PCHQR Program. At a minimum, if adopted, these measures should be excluded from public reporting, until such time as they have been revised to address the important considerations noted above. Possible New Quality Measure Topics for Future Years • • • • • Localized Prostate Cancer: Vitality Localized Prostate Cancer: Urinary Incontinence Localized Prostate Cancer: Urinary Frequency, Obstruction, and/or Irritation Localized Prostate Cancer: Sexual Function Localized Prostate Cancer: Bowel Function Memorial welcomes CMS ' s proposal to consider the inclusion of five patient-reported outcome measures for localized prostate cancer. We urge CMS to work with the ADCC, which has experience with implementing and reporting these measures, to incorporate these measures into the PCHQR program. • 30-Day Unplanned Readmissions for Cancer Patients Memorial Hospital for Cancer and Allied Diseases (Memorial) supports the adoption of the 30Day Unplanned Readmissions for Cancer Patients (3188) and believes it offers a significant step toward addressing the measurement gap in cancer care. Although reducing readmissions has been a CMS priority for over ten years, extant readmissions measures, such as the hospital-wide all-condition readmissions developed by Yale, exclude PPS-exempt cancer hospital patients by design given the complexity of their patient population. Until recently, no cancer-specific readmissions measure existed for benchmarking or to identify opportunities for improvement. The 30-Day Unplanned Readmissions for Cancer Patients measure takes into account important characteristics inherent in the nature of cancer care. For example, the measure uses "urgent" or "emergency" admission status to identify and focus on unplanned, potentially avoidable readmissions, where hospitals have a significant opportunity to identify interventions that may improve care. We agree it is appropriate to exclude readmissions from the numerator in which patients were discharged with an a priori intention to readmit for continued treatment, as is often the case with hematologic cancers. In addition, the measure excludes readmissions due to progression of disease, recognizing that complications associated with the disease itself(e.g., malignant ascites) are not preventable and are treated appropriately in the inpatient setting. Readmissions of patients with metastatic disease for reasons other than progression of disease are included, however, given that they may present opportunities to improve care. This approach represents a careful attempt to separate preventable from non-preventable readmissions in cancer patients. Ms. Seem Verma Centers for Medicare & Medicaid Services Department of Health and Human Services Page 5 of6 Risk adjustment that adequately reflects patient acuity and complexity is paramount in any cancer-specific measure. As noted in NQF's Readmissions 2017 Project report, the Technical Expert Panel raised concerns regarding the specific risk adjustment factors used in the 30-Day Unplanned Readmissions for Cancer Patients measure, particularly with regard to the use of "more than one comorbidity" as a single risk adjustment factor as well as "hospitalization in the prior 60 days," the latter of which may be a measure of quality itself. We believe the measure steward defended adequately their inclusion of these factors in the risk adjustment model, noting that the single comorbidity factor was included specifically as a means of addressing the wellknown problem of multi-collinearity among comorbidities, while literature has emerged supporting the notion that prior hospitalization serves as a proxy for severity of illness and care complexity. Our support notwithstanding, we note opportunities exist to further refine the risk adjustment model and look forward to future iterations of the measure. Accounting for Social Risk Factors in the PCHQR Program Memorial is supportive of implementing a well-designed approach to account for social risk factors in the PCHQR Program. To ensure that social risk factor adjustment supports its intended objective, i.e., a fair assessment of the quality of care, it will be necessary to review whether social risk adjustment is appropriate on a measure-by-measure basis. Individual SES/SDS risk factors should be applied only where there is a sound a priori rationale or empirical evidence demonstrating that the risk factor impacts measure performance in a manner that is partially or entirely outside the control of the health care provider and where the impact on measure performance is statistically significant. Where social risk adjustment is determined to be appropriate, a secondary analysis will be necessary to identify which social risk factors are both appropriate and available to be incorporated into the measure calculations without imposing additional administrative burden. This type of carefully applied social risk factor adjustment will help to ensure fairness and transparency in the PCHQR program and other CMS quality reporting programs. Extraordinary Circumstance Exceptions (ECE) Policy under the PCHQR Program Providing participants in the PCHQR Program with an extraordinary circumstance exception is an important mechanism to ensure that participants are not unduly burdened during natural disasters and other exigent circumstances. Memorial appreciates CMS 's proposal to extend an extraordinary circumstance extension to the PCHQR Program, and we endorse its adoption. Other Policy Changes: Other Operating Room (O.R.) and Non-O.R. Issues (41) Percutaneous Transfusion Memorial is concerned by CMS's proposal to reclassify the 20 ICD-10-PCS codes associated with inpatient percutaneous transfusion of bone marrow and stem cells to "non-O.R. procedures." While clinically correct, this reclassification would inappropriately assign some allogeneic transplants and nearly all autologous marrow and stem cell transplants into less Ms. Seem Verma Centers for Medicare & Medicaid Services Department of Health and Human Services Page 6 of6 specific, diagnosis-related DRGs, thereby disrupting the clinical homogeneity of the MS-DRGs to which they are newly assigned. An analysis by the American Society for Blood and Marrow Transplant found that, if implemented as proposed, less than 0.5 percent of autologous bone marrow and stem cell transplants in CMS's own data would be assigned to MS-DRG 016 (autologous bone marrow transplant with cc/mcc) and MS-DRG-017 (autologous bone marrow transplant without cc/mcc). Instead, these cases would be assigned to a wide range of less specific and less appropriate MS-DRGs. Failure to assign these patients to the appropriate transplant MS-DRGs will decrease the accuracy of all the approximately 70 MS-DRGs and associated payment levels affected by this reclassification; MS-DRGs to which transplants are reassigned will be artificially inflated, and the existing MS-DRGs for autologous transplant will be inaccurately categorized as low-volume. So, while it is clinically accurate to describe the 20 ICD-10-PCS codes identified as "Non-O.R. procedures," the grouper logic should be edited to ensure that patients receiving allogeneic transplants and autologous marrow and stem cell transplants continue to be assigned to MS-DRG 014, MS-DRG 016 and MS-DRG 017, respectively. Cell Acquisition Payment Policy Memorial urges CMS to reimburse the cell acquisition costs for allogeneic hematopoietic stem cell transplantation (HCT) in the same manner as it reimburses for the solid organ acquisition costs, i.e. , by reimbursing for the reasonable and necessary costs associated with acquiring the donor cells. The current reimbursement for allogeneic transplants under MS-DRG 014 does not account for acquisition costs, which average $48,436 for adult donor cells from marrow and peripheral blood stem cells and $65,117 from cord blood. Memorial urges CMS to amend its policy to extend reimbursement for the reasonable and necessary costs of acquisition to donor cells for allogeneic HCT. Again, we appreciate the opportunity to comment on the IPPS proposed rule for FY 2018. If you have any questions, please do not hesitate to call on us. Jorge Lopez, Jr.