Efficacy of advance directives in a general hospital.

UNLABELLED: A review of medical charts of all deaths for one year at a general acute care hospital reveals that 135/602 (22%) charts indicate that the patient had an advance directive. In 68/135 (50%) of the cases, the patients were unable to participate in decisions and met the conditions of the advance directive. In 33/68 (49%) of those cases the records indicate that the advance directive influenced care. In 63 of the 135 charts the advance directive was present and chart notations indicate an additional 25 advance directives were located at the physician's office. Eighteen of a total of 44 physicians listed as attending accounted for the 33 cases in which the record indicates that the advance directive was recognized. Twelve of these 135 patients were coded during their hospitalization. Three of the 12 were coded in the ER upon admission, the remaining 9 were coded in the course of their care in the acute care hospital. Regarding code status a three tiered (Cat I, II, III) classification system was in place. Initial classification of the 135 patients upon admission was: 64 "full code" (I), 56 were "all but CPR" (II), 15 were "No code" (III). Code classification at the time of death (or discharge) was: I = 45, II = 53, III = 36.

OBJECTIVE: To investigate the extent to which advance directives influence clinical care of patients during the final acute hospitalizations.

DESIGN: Retrospective chart review.

SETTING: General Hospital of 274 beds.

PATIENTS: 602 death charts reviewed, 135 contained indications or the execution of an advance directive.

MAIN OUTCOME MEASURES: The 1995 medical records of 602 death were reviewed for evidence of influence of advance directives in clinical care.

RESULTS: 24% of patients who had advance directives in the chart or at the physicians office had their directives recognized during their final hospitalization. In 68/135 (50%) of the cases the conditions to activate the advance directive were met. And in 33/68 (49%) of those cases the advanced directive was invoked. There was some, but less than expected correlation between advanced directives and DNR orders. In a three tiered Code Category Classification system (Cat. I, full code, Cat. II Chemical Code, Cat. III, No Code.) the initial classifications in the 135 cases with evidence of advance directives were Cat. I 47%, Cat. II 42%, and Cat. III 11%. Compared to 59 cases where there was no indication of an advance directive the classifications were Cat. I 67%, Cat. II 26% and Cat. III 7%. However, the classifications in the two groups at the time of death of the patients were Cat. I 34% & 31%, Cat. II 39% & 39% and Cat. III 27% & 30%. There was a 20% increased incidence of an initial classification of full code in the cases without indication of an advance directive. But once the patient care involved review of code status, the final classifications of patients were the same irrespective of the presence of an advance directive.

CONCLUSIONS: In 50% or 68/135 of the cases the patient met the conditions for invocation of the advance directive and in 33 or 49% of those cases the advance directive was invoked. Another way to state the impact of advance directives in the population studied is that in 22% of the 602 deaths there was indication of an advance directive and in 50% of those cases the directive became relevant and in 49% of those cases it had a bearing on the care (or in 5% of the 602 death studied). More research is needed to determine why advance directives are not utilized more and why they to do not have greater effect on clinical care decisions in terminal patients. But problems with making them available to relevant parties, hospital record keeping, and physician recognition of their significance are evident.