Effect of guidelines on management of head injury on record keeping and decision making in accident and emergency departments.

Abstract

OBJECTIVE--To compare record keeping and decision making in accident and emergency departments before and after distribution of guidelines on head injury management as indices of implementation. DESIGN--Before (1987) and after (1990) study of accident and emergency medical records. SETTING--Two accident and emergency departments in England. PATIENTS--1144 adult patients with head injury in department 1 (533 in 1987, 613 in 1990) and 734 in department 2 (370, 364 respectively). MAIN MEASURES--Recording of relevant symptoms and signs as determined in the guidelines; presence of, indications for, and rates and appropriateness of skull x ray examination and admission. RESULTS--The median number of guidelines variables recorded for all study periods ranged from 7 to 9 out of a possible maximum of 27. For key decision making variables the presence or absence of penetrating injury was least likely to be recorded (< or = 1%) and that of loss of consciousness most likely (> or = 75%). Altogether, the proportion of patients receiving skull x ray examination or admitted varied from 25%-60% and 7%-23% respectively; overall, 69% (1280/1856) and 64% (1177/1851) of patients were managed appropriately. However, no consistent change occurred in the departments between the study periods. For instance, in department 1 the proportion of appropriate x ray examinations rose significantly after distribution of the guidelines (from 61% (202/330) to 73% (305/417)) and appropriate decisions on whether to x ray or not also rose (from 65% (340/522) to 72% (435/608)). There was no significant change in department 2, although the proportion of appropriate admissions fell (from 33% (55/166) to 15% (19/130)). CONCLUSIONS--Recording practice and decision making were variable and had not consistently improved after dissemination of the guidelines. Strategies are required to ensure effective implementation of guidelines.