Death Among Patients Hospitalized With Pneumonia: Implications for Hospital Outcome Measures

The Affordable Care Act established the Value-Based Purchasing Program, launched in 2013, which uses risk-standardized mortality rates as a benchmark to penalize or reward hospitals.¹ The risk-standardized mortality rate is calculated using administrative data and includes patients with life-threatening illnesses for whom pneumonia may be the last insult in the setting of terminal illness. Moreover, preferences for limiting the use of or withdrawing life-sustaining therapy are considered only for patients already enrolled in hospice on the day of admission. Therefore, risk-standardized mortality rates measure the outcome of care but not necessarily the appropriateness and value of care.²

We sought to determine the proportion of patients identified with pneumonia by the Centers for Medicare & Medicaid Services risk-standardized mortality measures for whom pneumonia was a major contributor to death and to describe the intensity of care and patient preference for life-sustaining therapies.

Centers for Medicare & Medicaid Services criteria³ were used to identify all adult patients who died with a principal diagnosis of pneumonia between January 1, 2008, and December 31, 2012, at 3 Massachusetts hospitals.

Guided by the Mortensen et al⁴ classification schema, 2 of us (R.J. and J.F.) assessed patients’ medical records to determine if pneumonia was a minor or major contributor to death. Pneumonia was considered a major contributor if the patient had stable medical conditions and death would not have occurred in the absence of pneumonia, and a minor contributor if the patient had advanced life-threatening illnesses (ie, met criteria for palliative care)⁵ and pneumonia was on the final pathway to death.

The study was approved by the Baystate Medical Center Institutional Review Board. As this was a retrospective chart review, no patient consent was necessary.

A total of 202 deaths were included; mean patient age was 78.5 years, 54.5% of patients were female, and 56.4% had a do-not-resuscitate order at admission. During hospitalization, 30.2% were admitted to an intensive care unit, 23.8% were intubated, and 24.8% died in the intensive care unit (Table 1).

Most patients had severe debilitating illnesses: 24.1% had advanced dementia, 9.3% showed failure to thrive, 18.2% had cerebrovascular disease with severe functional impairment, and 7.4% had lung cancer. In addition, 2.9% of patients had a feeding tube and 1.9% received long-term mechanical ventilation.

Pneumonia played a major role in the deaths of 37 patients (18.3%). Examples of deaths with pneumonia as a major and minor contributor appear in Table 2. Compared with patients with pneumonia as a minor contributor, patients with pneumonia as a major contributor received more intense care. Of 165 patients with life-threatening illnesses, 57.6% had do-not-resuscitate orders at admission and 57.0% refused intubation. Invasive and noninvasive mechanical ventilation were discontinued before death in 83.3% and 91.2% of the patients with life-threatening illnesses, respectively. Of the 202 deaths, 95 patients (47.0%) had life-limiting illnesses meeting the criteria for palliative care and had do-not-resuscitate orders at admission.

In this detailed retrospective medical record review of patients identified with pneumonia by the Centers for Medicare & Medicaid Services risk-standardized mortality rate measures, we found that pneumonia was a major contributor to death in only 18.3% of cases. Almost half of the deaths occurred among patients who, at the time of admission, had appropriately decided to forgo aggressive treatment. The deaths of these patients cannot be assumed to represent poor-quality care because survival was not necessarily the goal of therapy. In many other cases, care was ultimately withdrawn, but we were unable to determine whether the overall quality of care contributed to the patient’s death.

Only 57.6% of the patients with advanced illnesses had do-not-resuscitate orders and many of these patients received aggressive care, which suggests opportunities to improve end-of-life discussions. Currently, the mortality measures include patients with a terminal illness and may penalize hospitals that take a more patient-centered approach and use palliative care, while encouraging hospitals to provide inappropriately aggressive treatment when a patient is at the end of life.^2,6

The findings of this study suggest that mortality measures could be enhanced by taking into account patient preferences for treatment and end-of-life care.

Corresponding Author: Mihaela S. Stefan, MD, Department of Medicine, Baystate Medical Center, 759 Chestnut St, Springfield, MA 01199 (mihaela.stefan@baystatehealth.org).

Published Online: March 16, 2015. doi:10.1001/jamainternmed.2015.114.

Author Contributions: Dr Stefan had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: All authors.

Acquisition, analysis, or interpretation of data: Stefan, Jaber, Garb, Fitzgerald, Rothberg.

Drafting of the manuscript: Stefan, Jaber, Garb.

Critical revision of the manuscript for important intellectual content: Stefan, Lindenauer, Garb, Fitzgerald, Rothberg.

Statstical analysis: Stefan, Garb.

Administrative, technical, or material support: Jaber, Fitzgerald.

Study supervision: Jaber, Lindenauer, Fitzgerald, Rothberg.

Conflict of Interest Disclosures: Dr Lindenauer receives support to develop hospital outcome measures from the Centers for Medicare & Medicaid Services. No other conflicts were reported.

Funding/Support: Dr Stefan is supported by grant 1K01HL114631-01A1 from the National Heart, Lung, and Blood Institute of the National Institutes of Health and by grant UL1RR025752 from the National Center for Research Resources and the National Center for Advancing Translational Sciences, National Institutes of Health.

Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The content of this article is solely the responsibility of the authors and does not represent the official views of the National Institutes of Health.

Additional Contributions: Anu Joshi, Clinical Research Coordinator, Center for Quality of Care Research, Baystate Medical Center, provided assistance in editing the manuscript and preparing the tables. Ms Joshi was not compensated for her contribution.