Can extubation time be predicted?
© Current Science Ltd 1999
Received: 26 November 1999
Published: 23 December 1999
Fast-tracking has become an established practice in many cardiac units. We have attempted to identify the factors that may result in prolonged ventilation in our unit.
Materials and methods
The audit was conducted prospectively over an 8-week period during February and March 1999. The audit forms of a total of 156 patients who passed through the intensive care unit were analyzed. Appropriate criteria for extubation were set out beforehand, as were guidelines for management of pain, shivering, agitation and confusion, and poor gas exchange. The following data was collected: patient details; operation, anaesthetist and surgeon; anaesthetic technique; preoperative risk stratification (Parsonnet and Euro-scores); time to extubation; principal reason for continued ventilation (recorded every 2 h for 24 h); and hospital discharge date.
The patients were divided into four principal groups according to their extubation times: less than 6 h, 6–12 h, 12–24 h and longer than 24 h. We focused particularly on the group of patients with extubation times less than 6 h and found that 25.7% were extubated during this period (as compared with only 3% from previous audit figures in 1997). Neither the Parsonnet or Euroscore predicted extubation times well. The percentage of patients extubated in under 6 h divided into low, medium and high risk groups for Parsonnet were 29.3, 15.15 and 34.6%, and for Euroscore 26.5, 29.3 and 16.7%. Very high scores were associated with longer ventilation times as expected. The percentage of patients in the less than 6 h group was found to vary between individual anaesthetists from 8.0 to 50% and surgeons 7.7–31.6%. This degree of variation could not be explained by preoperative risk stratification alone. Patients first on the theatre list (am) were more likely to be extubated within 6 h (32.1%) compared with patients who were later on the list (pm; 18.2%; P < 0.05). This could not be explained by risk stratification. Of 'am patients' 74.1% were discharged by day 6 compared with 60% of 'pm patients' (P = 0.08). When we examined the effects of coexistent disease we found that smoking had little effect (25.8% <6 h), presence of chronic obstructive pulmonary disease delayed extubation (16.7%) and excessive alcohol intake appeared to expedite extubation (37.5%) compared with all patients (25.6%), but these associations did not reach statistical significance. The drugs used to provide hypnosis and analgesia on cardiopulmonary bypass were recorded, and the percentage of patients extubated in under 6 h were as follows: midazolam 13.7%, volatile on bypass 16.1%, propofol infusion 30.3% and methohexitone 44.4%, remifentanil 14.3%, fentanyl 25.0% and alfentanil 38.5%. The common causes of continued ventilation at 6 h were 'too drowsy' (70%) and 'poor gases' (28%), but at 24 h were 'inotropic support' (86%), 'confusion' (43%) and the use of 'intra-aortic balloon pump' (37%) were cited by the attending nurse. The percentages of patients discharged by the sixth postoperative day were 80.0% in the less than 6 h group, 70.6% in the 6–12 h group, 57.6% in the 12–24 h group and 0.0% in the longer than 24 h group (P < 0.05).
Early extubation may expedite recovery and hospital discharge. Extubation time was influenced by both anaesthetist and the anaesthetic technique used. Current preoperative risk scoring systems are insensitive predictors of ventilation times. Patients that had their operations later in the day were more likely to be ventilated overnight, perhaps unnecessarily. Clinical audit changes practice and can be a very useful tool.