Mortality is the ultimate outcome metric for patient care. In the hospital, measuring mortality appears simple and straightforward, but upon considering the layers of attribution and risk stratification, it becomes significantly more complex. Measures of mortality are used in various quality reporting and national ranking programs, making them a common focus of hospital leadership. However, the connection between hospitalists and measures of mortality is not always clear.
Case
Mrs. Smith is an 86-year-old woman who presented to the hospital with abdominal pain. Initial evaluation showed acute cholecystitis, and surgery was consulted for a laparoscopic cholecystectomy. Medicine was also consulted due to her underlying hypertension, diabetes mellitus type 2, hyperlipidemia, and coronary artery disease. The operation was completed successfully, but she developed post-operative pneumonia. She was transferred to the hospital medicine team for treatment. Unfortunately, the patient’s respiratory status worsened and required transfer to the intensive care unit. In the unit, her condition worsened further, and goals of care conversations with the family led to hospice enrollment. The patient was discharged to the hospice unit and passed the next day.
- Should this mortality be connected to the hospital encounter?
- Would the answer to the above question change if the patient was deemed hospice appropriate, but the family chose not to enroll?
- Which service should the patient be linked to (surgery, hospital medicine, intensive care unit)?
- Was this mortality expected based on their condition?
A variety of names are used for measures of mortality, including mortality rate, mortality index, and ratio of observed to expected, or O/E, mortality. It is often initially assessed at the level of the hospital, but it can also be evaluated by specific disease states or specialty service lines within a hospital, or in a broader context of an accountable care organization or health plan. Institutions or organizations may choose to attribute at an even more granular level, such as groups, specific hospital service lines, or potentially at the individual level.
Appropriate attribution can be difficult given the complex nature of hospital-based care, especially if taken to the individual hospitalist level, as illustrated in the case above. The level of contribution any one provider makes to a patient’s ultimate outcome is difficult to ascertain, as the entire continuum of care is involved in a patient’s clinical trajectory before, during, and after a hospitalization that can prevent or lead to a patient’s passing. Primary care plays a significant role in this measure as well, as they manage the patient’s chronic diseases and ideally start advanced care planning, especially with known severe or terminal conditions.
Mortality measures are typically reported at monthly, quarterly, or yearly intervals. At the most basic level, the data are easy to obtain for the number of discharges alive compared to the number of discharges deceased. Risk-adjusted mortality rates require more information to calculate and can go through complex formulas managed by third-party analytics organizations, such as Vizient, Midas, or Premier. The addition of risk adjustment adds an opportunity to improve the outcome from a very different approach. Unadjusted mortality focuses on the care provided, and improvement opportunities include clinical process measures, such as guideline-directed therapy for heart failure or meeting sepsis time-to-treatment goals.
Risk stratification is based on the documented primary diagnosis for the patient, in addition to other clinical conditions that add to the patient’s complexity. The primary focus of improvement in this area is provider documentation, which can be a robust enterprise within many programs. Documentation offers an opportunity for a hospitalist to affect mortality metrics on a case-by-case basis, with an aggregated improvement in documented severity of illness. There are examples of institutions improving their reported risk-adjusted mortality through the implementation of targeted documentation-improvement programs.1,2
Other opportunities to improve reported mortality metrics can be dependent on the specific definitions for mortalities attributed to a hospitalization. One common definition excludes patients who passed while in the care of a hospice, even if directly enrolled from the hospitalization. This could create pressure to enroll patients in hospice who are imminently passing and were raised publicly as a concern.3 A balancing measure for this potential unintended consequence is measuring time from hospice enrollment to passing. Another exclusion in some programs is patients being in observation status rather than inpatient at the time of passing. The use of observation status has been described in specific population workflows, but has not been more universally evaluated.4 In other measure models, hospitals are accountable for the patient’s survival for 30 days following discharge, which includes both public reporting in Hospital Compare and, in some cases, payment adjustments from the Centers for Medicare and Medicaid Services.5 A greater emphasis on connection post-discharge care is required for improvement in these measures.
Conclusion
With all the complexities of risk adjustments and multiple definitions, it can be easy for a frontline hospitalist to get lost in how they can contribute to improvement. Focusing on mortality can also generate emotional strain on hospitalists, as many of the patients passing on hospitalist services pass from acute events, such as codes. Attribution at the group or service line level can help to avoid some of the individual strain, while keeping the measure relevant. The case above and related questions highlight how attribution, mortality definitions, and multi-disciplinary care go into measuring mortality. Improvement efforts within hospitalist groups can range from documentation improvement and hitting sepsis targets to enhanced advanced-care planning. The strategy that best fits a group will depend on what definition their organization prioritizes and the related resources available.
Click here for more information on SHM’s Quality Measures.
Dr. Cerasale
Dr. Gershfield
Dr. Virapongse
Dr. Talari
The authors are members of SHM’s Performance Measurement and Reporting Committee, which created this series to explore quality measures common in hospital medicine. Dr. Cerasale is the outcomes quality director for UChicago Medicine, quality improvement director for the section of hospital medicine, and associate professor of internal medicine residency at the University of Chicago in Chicago. Dr. Gershfield is a hospitalist at Sequoia Hospital in Redwood City, Calif., and director of quality and performance for hospital medicine at Vituity. She is also a member of The Hospitalist’s editorial board. Dr. Virapongse is an associate vice-chair for quality in the department of medicine, the director of quality, patient safety, and experience in the division of hospital medicine, and an associate professor at the University of Colorado School of Medicine in Aurora, Colo. Dr. Talari is a hospitalist at Ballad Health.
References
1. Kessler BA, et al. Reducing the reported mortality index within a neurocritical care unit through documentation and coding accuracy. World Neurosurg. 2020;133:e819-e827. doi: 10.1016/j. wneu.2019.10.022.
2. Horwood CR, et al. Improving the mortality index by capturing patient acuity through interprofessional real-time documentation improvement in a single hospital system. Surgery. 2018;164(4):687-693. doi: 10.1016/j.surg.2018.04.045.
3. Morgenson G. ‘You’re not God’: Doctors and patient families say HCA hospitals push hospice care. NBC News website. https://www.nbcnews.com/health/healthcare/doctors-say-hca-hospitals-push-patients-hospice-care-rcna81599. Published June 21, 2023. Accessed January 10, 2026.
4. Baugh CW, et al. A hospice transitions program for patients in the emergency department. JAMA Netw Open. 2024;7(7):e2420695. doi: 10.1001/jamanetworkopen.2024.20695.
5. Center for Medicare & Medicaid Services. Hospital Readmissions Reduction Program. CMS.gov website. https://www.cms.gov/medicare/quality/value-based-programs/hospital-readmissions. Accessed January 12, 2026.