DNR Orders May Skew Hospital Mortality Rates and Rankings

1. Based on a retrospective review of pneumonia admissions in California and without accounting for Do-not-resuscitate (DNR) rates, hospitals with higher DNR rates had higher mortality rates. However, after stratifying for patient DNR status, hospitals with higher DNR rates had better hospital survival, both for patients with DNR and non-DNR status.

2. Additionally, hospitals with higher DNR rates were more likely to discharge patients home.
3. These results demonstrate the importance of accounting for DNR rates in hospital quality measures and rankings, as failure to do so may penalize hospitals that treat a higher proportion of patients who mandate limits on their care measures.

DNR orders may skew hospital mortality rates and rankings. Image PD

Evidence Rating Level: 2 (Good)

Study Rundown: DNR orders convey a patient’s wish to forego cardiopulmonary resuscitation in the event of a cardiac arrest. Early DNR orders (within 24 hours of hospitalization) are likely to indicate a patient’s pre-illness wishes based on their comorbid conditions and family discussions. However, these DNR orders are not currently included in measuring hospital outcomes, even though they could indicate a worse prognosis. As a result, this study, which examined pneumonia admissions in the state of California during one calendar year, was conducted to determine whether accounting for DNR status affected hospital mortality and other outcomes. The results of the study showed that patients with early DNR orders had a higher risk of death. However, after stratifying for patient DNR status, hospitals with higher DNR rates had better hospital survival, both for patients with DNR and non-DNR status. Additionally, hospitals with higher DNR rates were more likely to discharge patients home and had shorter length of stay.

A key strength of the study was the relatively open inclusion criteria in a large state, which is representative of the US. A notable weakness was that in-hospital survival may not be the most useful marker of improved outcomes, since a significant portion of patients who were DNR may have higher, albeit expected, out of hospital mortality after their incident hospitalization.

In-Depth [retrospective cohort]: The cohort for this study, obtained from Healthcare Cost and Utilization Project California State Inpatient Database (CA SID) consisted of adults 40 years and older hospitalized for pneumonia (either as a primary or secondary diagnosis) during 2011. Early DNR was defined as a DNR order within the first 24 hours of hospitalization. Hospital DNR rate was the percentage of patients in this cohort who had an early DNR order at each hospital examined. The primary outcomes were association between early DNR orders and hospital DNR rates with in-hospital mortality. Also, the association between hospital DNR rates and discharge destination and length of stay amongst survivors was examined. In the final analysis, there were 90,644 pneumonia cases among 303 California hospitals identified. The results revealed higher mortality amongst the early DNR patients (OR, 3.74; 95% CI, 3.55