| Summary: | 碩士 === 國立臺灣大學 === 公共衛生碩士學位學程 === 105 === Abstract
Background: As the impact of intraoperative undesired incident events on the anesthesia-related adverse outcomes is not well-addressed, it is valuable to elucidate whether and what kinds of intraoperative undesired events are associated with anesthesia-related adverse outcomes. Beside, most of studies treat a constellation of outcomes with repeat to postoperative complications as independent one in the same individual patients, while studying the relationship of undesired incident events to postoperative complications. The failure of considering these multiple outcomes as correlated outcome may affect the effect sizes regarding the association between intraoperative undesired incident events and postoperative complications. As factors in association with adverse events are numerous how to clarify which factors are medicated through another set of factors is of great interest. By using anesthesia clinical outcomes registry data from one medical center in Taiwan, we have chance to apply a series of statistical models making allowance of correlation property in the association study between intraoperative undesired event and postoperative complications
Aims: The aims of this thesis were
1.to explore the incidence of intraoperative undesired incidents and postoperative anesthesia adverse events;
2.to elucidate whether the intraoperative incidents have influence on the postoperative adverse events by using proportional odds regression models;
3.to investigate the effect of the intraoperative undesired incidents on multiple postoperative adverse events taking into account the correlation property within subject by using general estimating equation model;
4.to assess the anesthesia risk factors with proportional odds regression models on the newly classified risk groups;
5.to investigate the conservative relationships between postoperative adverse events, resulting in postoperative complications by using confirmatory factor analysis;
6.to tailor the clinicians to anesthesia decision to reduce the occurrence of these side effects and to ensure the quality assurance of care.
Material and method:By using a prospective observation study design, a cohort of patients receiving operation at Taipei or Tamsui branch of Mackay Memorial Hospital between 2011 and 2014 were enrolled. The quality indices including 46 items and eight aspects considering the intraoperative undesired incidents/postoperative adverse events were collected. The 14 defined outcomes of undesired events were ascertained by an interview performed by trained anesthetic registers within 24 hours after the operation. Total of 88,990 patients were enrolled in our analysis.
A series of conventional proportional odds regression models were applied to assess the significance between the items of the intraoperative undesired incidents and each of the 14 undesirable events adjusting for age, sex, and anesthesia techniques. This is followed by applying the generalized estimating equation models to take into account the correlated characteristics of the 14 undesirable events when assessing the significance of the multiple quality indices. Alternatively, we reclassified patients into different groups according to the summation of the 14 adverse events for a better reorganization of the risk factors associated with the spectrum of postoperative adverse events. We further applied confirmatory factor analysis to elucidate the influence of the intraoperative undesired incidents on the multiple adverse events with the consideration of paths between five components toward adverse events (tissue injury, vital system injury, symptomatic response, respiratory response, and vital signs).
Results:A total of 88,990 patients underwent surgery in MacKay Memorial Hospital were recruited in the current analysis. After adjustment for patient factors, there were strong relation between postop adverse events and periop anesthesia-related incidents. Gender difference was not consistent across postoperative complications. Male had higher risk for hemodynamic instability (OR=1.28, 95% CI: 1.05, 1.57), conscious disturbance (OR=1.75, 95% CI: 1.27, 2.40), upper airway discomfort or dyspnea (OR=1.66, 95% CI: 1.45, 1.90), respiratory assistant devices requirement (OR=1.55, 95% CI: 1.43, 1.69), and abnormal thermoregulation (OR=1.29, 95% CI: 1.19, 1.39), whereas female were more likely to have reactions of nausea (OR=0.45, 95% CI: 0.42, 0.48), vomiting (OR=0.36, 95% CI: 0.34, 0.39), and dizziness (OR=0.52, 95% CI: 0.49, 0.54) after surgery. Age, and anesthesia techniques were consistent risk factors. Compared with sedation, general anesthesia had 2.33-19.94 fold risk for most postoperative adverse events, except 530-fold for respiratory abnormality, and regional anesthesia had smaller effects with a range of 1.22-9.77. ASA class III+ was more likely to have postoperative complications, except nausea, dizziness, and peripheral nerve injury, with which ASA class III+ was negatively associated.
When we examined the intraoperative undesired events, none under the categories of preoperative preparation and nervous system was significant associated with postoperative complications, except inadequate depth of anesthesia/awake during operation which was strongly associated with perioperative awareness (OR=267, 95% CI: 57-1248) (but only affecting this complication). In the realm of perioperative process, unplanned ICU admission was the most significant one. In the realm of medicine and transfusion, blood transfusion with more than four units was the most important one. For respiratory system, hypoxia, hypercapnia, aspiration of stomach contents were commonly associated with complications after operation. The detailed items listed in the cardiovascular system commonly contributed to postoperative complications. Hyperthermia and oliguria were two major intraoperative undesired incidents which contributed the largest number of postoperative complications.
When the multiple postoperative complications were considered simultaneously, female (OR=1.40, 95% CI: 1.37, 1.45), elderly age with increasing risk by 1% per 1 year of age, ASA class III+ (OR=1.34, 95% CI: 1.29, 1.39), and general (OR=9.07, 95% CI: 8.44, 9.75) or regional (OR=5.28, 95% CI: 4.88, 5.71) anesthesia were identified as risk factors. unplanned ICU admission (OR=1.49, 95% CI: 1.14, 1.94), blood transfusion with more than four units (OR=1.70, 95% CI: 1.59, 1.83), cardiac arrest (OR=2.24, 95% CI: 1.16, 4.32), tachycardia requirement (OR=1.43, 95% CI: 1.22, 1.67), hypertension (OR=1.12, 95% CI: 1.01, 1.25), hypotension (OR=1.15, 95% CI: 1.09, 1.22), complications of central nerve system (OR=4.21, 95% CI: 4.12, 4.31), injury of teeth (OR=1.89, 95% CI: 1.23, 2.90), lip/oral cavity/nose (OR=1.36, 95% CI: 1.07, 1.72), skin (OR=1.83, 95% CI: 1.42, 2.34), and eyes (OR=2.10, 95% CI: 1.29, 3.42), hypothermia (OR=1.24, 95% CI: 1.13, 1.35), hyperthermia (OR=1.53, 95% CI: 1.22, 1.92), oliguria (OR=1.57, 95% CI: 1.41, 1.75), and severe vomiting (OR=2.33, 95% CI: 1.68, 3.23) were more likely to have postoperative complications. The findings for collective categories of intraoperative events were similar to previous findings. The consistent findings were also found when the outcome of interests was the level of numbers of postoperative complications.
Using the confirmatory factor analysis to assessing the relevance from the multiple items of quality indices to the undesired events and thereafter between the cluster of undesired events, the significant path through the component of symptomatic response to that of abnormality in vital signs (0.06 (SE=0.007, t=8.13)) and that from respiratory response to that of abnormality in vital signs (0.56 (SE=0.026, t=21.10)) were revealed. The significance between the predictor of quality indices to the undesired events and that of each component measured by relevant undesired events were also found.
Conclusions:Age, gender, ASA class, and anesthesia techniques were associated with postoperative adverse events. Intraoperative undesired incidents in medicine/transfusion, respiratory system, cardiovascular system, and other ungrouped factors are four major undesired incident events that account for postoperative adverse events or adverse risk scores. These risk factors making different contribution to different categorized adverse events were noted. The results of confirmatory factor analysis can be applied to aiding in clarifying the path of adverse events in clinical practice.
|
|---|
