The author replies

    loading  Checking for direct PDF access through Ovid

Excerpt

We recently demonstrated that automated early warning system (EWS) alerts could identify hospitalized patients potentially benefitting from palliative care discussions (1). Furthermore, this intervention was associated with significantly more patients having their resuscitation preferences and advance directives documented in the medical record as well as being associated with fewer transfers to the ICU. In his letter, Dr. Chertoff (2) emphasizes that prior research has shown that race influences the use of life-sustaining treatments (i.e., invasive mechanical ventilation, hemodialysis, and feeding tubes) in hospitalized patients with African-American patients being less likely to receive comfort care when compared with white patients. Indeed, we previously examined this issue in this journal identifying three independent predictors for the withdrawal of life-sustaining interventions (3). Having a planned therapeutic trial of life-sustaining interventions outlined in the medical record, lack of a private attending physician and the presence of clearly defined advance directives regarding patient preferences for medical care were associated with a greater likelihood of having withdrawal of life-sustaining interventions. There was no overall difference in the withdrawal of life-sustaining interventions according to race in that study (3). However, African-American patients were significantly less likely to have a private attending physician and significantly less likely to have private health insurance when compared with white patients. Dr. Chertoff (2) speculated on whether automated EWS alerts would be equally effective at stimulating palliative care discussions in African-American patients and white patients and asked us to perform a post hoc analysis addressing this point.
Table 1 shows the results of our post hoc analysis assessing the 202 African-American and white patients in our trial. African-Americans were significantly younger, had similar comorbidities (data not shown), but were less likely to have advance directives documented or resuscitation status documented prior to generating an EWS alert during their hospital stay. A first request for advance directives, documentation of resuscitation status prior to hospital discharge, a documented change in resuscitation status, and delivery of a formal palliative care consultation was similar for both races. However, having documented advance directives prior to hospital discharge was significantly less common among African-American patients. We also found that hospital mortality and ICU transfer were significantly greater for white patients, whereas hospital length of stay was less for African-American patients. One of the limitations of our study is that severity of illness may not have been adequately characterized (we employed age and comorbidities), which may have accounted for some of the outcome differences observed.
In summary, our post hoc analysis would suggest that racial differences in documentation of advance directives can vary by race among hospitalized patients generating an EWS alert, with most of this disparity being present prior to generation of the EWS alert. Furthermore, the delivery of specific interventions linked to palliative care (request for first-time advance directives, formal palliative care consultation, and change in resuscitation status) did not vary by race. The impact of socioeconomic factors on palliative care for hospitalized patients seems to be complex and dependent on many factors. We agree with Drs. Huber and Edelson (4) that more research is needed to determine if automated systems can overcome these factors in order to provide more uniform palliative care to those who might benefit from it (5).
    loading  Loading Related Articles