Risk Factors for Delirium Following Cardiac Surgery

Overview

This study will evaluate several potential risk factors for postoperative delirium in cardiac surgery patients. The risk factors will include use of specific preoperative, intraoperative, and/or postoperative medications. Other risk factors investigated may include exposure to cardiopulmonary bypass, surgical technique, and or duration of surgery.

In order to understand the role of medication usage and/or cardiopulmonary bypass, the investigators propose to leverage the perioperative data warehouse (PDW), an IRB approved data registry, to retrospectively evaluate the subjects. All patient data available in the warehouse may be included in the review. At this time, there are approximately 750,000 patients in the PDW. All data will be accessed retrospectively after the patient has been seen for their care here at Vanderbilt. The Vanderbilt Perioperative Information Management System (VPIMS) is part of an integrated system that covers the entire perioperative process including instrument \& supply management, scheduling, and status displays. Because of the extensive effort that has been put toward in-house intraoperative software development, the Vanderbilt Anesthesiology \& Perioperative Informatics Research (VAPIR) Division is able to benefit from the availability of thousands of historical intraoperative records, each containing detailed physiologic information. The PDW incorporates VPIMS and electronic healthcare records (EHR) to build its database. In addition to PDW, information from the Society for Thoracic Surgeons (STS) database will be used in our study. The database uses algorithms created by programmers with the assistance clinicians to screen the data and assess for accuracy and completeness. Once the data is compounded and uploaded, each variable is screened for outliers by a clinician. Outliers will be assessed and validated by a clinician based on trends from the raw data. The data collected for this study includes an extensive data dictionary listed below: variable teeid anescasenumber mrn date orintime orouttime age gender race hx\_dialysis\_vpec hx\_dialysis\_hemodialysis\_vpec hx\_dialysis\_peritoneal\_vpec hx\_esrd creatinine\_base creatinine\_pod1 creatinine\_pod2 creatinine\_pod3 creatinine\_pod4 creatinine\_pod5 creatinine\_pod6 creatinine\_pod7 creatinine\_pod8 creatinine\_pod9 creatinine\_pod10 hx\_htn hx\_mi ck\_preop ckmb\_preop troponin\_preop hx\_cabg hx\_angioplasty hx\_angina hx\_chf hx\_ef\_less\_35 hx\_chf\_recent\_hospital hx\_chf\_stable hx\_chf\_currentsigns hx\_afib hx\_aflutter hx\_carotid\_disease\_current hx\_carotid\_disease\_past hx\_dementia hx\_depression hx\_stroke hx\_tia hx\_claudication chol\_base ldl\_base hdl\_base triglycerides\_base hx\_copd hx\_emphysema hx\_smoking\_current hx\_smoking\_quitless3months hx\_smoking\_quitmore3months hx\_osa hx\_hypoxia\_ra\_sat\_less92 hx\_home\_o2 diabetes hba1c charlson\_index weight height bmi hct\_base hb\_base platelet\_count\_base hx\_neoplasm statin\_preopyesno clopidogrel\_preop aspirin\_preop betablocker\_preop aceinhibitor\_preop diuretic\_preop preop\_meds ph\_intraop\_first ph\_intraop\_last ph\_intraop\_min ph\_intraop\_max ph\_intraop\_median ph\_intraop\_count po2\_intraop\_first po2\_intraop\_last po2\_intraop\_min po2\_intraop\_max po2\_intraop\_median po2\_intraop\_count lactate\_intraop\_first lactate\_intraop\_last lactate\_intraop\_min lactate\_intraop\_max lactate\_intraop\_median lactate\_intraop\_count ebl\_intraop hespan\_intraop lr\_intraop lr5dex\_intraop ns\_intraop plasmalyte\_intraop input\_intraop uop\_intraop emergency\_surgery procedure\_cpt procedure\_descriptor cabg\_grafts cpb\_on\_time1 cpb\_off\_time1 cpb\_on\_time2 cpb\_off\_time2 cpb\_on\_time3 cpb\_off\_time3 cpb\_on\_time4 cpb\_off\_time4 cpb\_on\_time5 cpb\_off\_time5 cpb\_duration cpb\_duration\_calculated cbp\_recorded\_matches\_calculated ao\_clamp\_duration circ\_arrest\_duration ultrafiltration\_volume pump\_drugs uop\_onpump prbc\_intraop\_ml prbc\_intraop\_unit ffp\_intraop\_ml ffp\_intraop\_unit platelets\_intraop\_ml platelets\_intraop\_unit cryo\_intraop\_ml cryo\_intraop\_unit prbc\_postop ffp\_postop platelets\_postop cryo\_postop ckmb\_pod1 req\_reop\_24 req\_reop\_72 is\_reop wbc\_base wbc\_pod1 wbc\_pod2 wbc\_pod3 wbc\_pod4 wbc\_pod5 wbc\_pod6 wbc\_pod7 wbc\_pod8 wbc\_pod9 wbc\_pod10 ph\_pod1 ph\_pod2 ph\_pod3 ph\_pod4 ph\_pod5 ph\_pod6 ph\_pod7 ph\_pod8 ph\_pod9 ph\_pod10 lactate\_pod1 lactate\_pod2 lactate\_pod3 lactate\_pod4 lactate\_pod5 lactate\_pod6 lactate\_pod7 lactate\_pod8 lactate\_pod9 lactate\_pod10 pfratio\_pod1 pfratio\_pod2 pfratio\_pod3 pfratio\_pod4 pfratio\_pod5 pfratio\_pod6 pfratio\_pod7 pfratio\_pod8 pfratio\_pod9 pfratio\_pod10 creatinine\_pod30 creatinine\_pod60 creatinine\_pod90 creatinine\_pod365 dialysis\_postop\_predischarge dialysis\_postop\_pod30 dialysis\_postop\_pod60 dialysis\_postop\_pod90 dialysis\_postop\_pod365 cam\_icu\_pod1 cam\_icu\_pod2 cam\_icu\_pod3 cam\_icu\_pod4 cam\_icu\_pod5 cam\_icu\_pod6 cam\_icu\_pod7 cam\_icu\_pod8 cam\_icu\_pod9 cam\_icu\_pod10 rass\_pod1 rass\_pod2 rass\_pod3 rass\_pod4 rass\_pod5 rass\_pod6 rass\_pod7 rass\_pod8 rass\_pod9 rass\_pod10 los\_icu los los\_surgery\_discharge death\_inhospital death\_date death\_pod30 death\_pod60 death\_pod90 death\_pod365 afib\_postop afib\_postop\_datetime afib\_hours\_between\_orouttime\_and\_afib\_postop\_datetime afib\_postop\_ecg stroke\_postop ventilation\_time The size of the registry for cardiac surgery patients in currently over 8,000 patients which should be more than enough to demonstrate an effect. The following is a brief description of the investigators data analysis: * Categorical variables will be expressed as frequency and percentage. * Continuous variables will be reported as median and interquartile range (IQR). * the investigators will compare the prevalence of medication use with the incidence of postoperative delirium and adjust for confounders. * Analysis will initially be completed with adjusting for confounders, and then consider matching patients with propensity score based on initial results. * Statistical significance (p≤0.05) for categorical variables will be computed using the Fisher's exact test or chi-squared and the Deuchler-Wilcoxon procedure or t-test for continuous variables. * An iterative expectation-maximization (EM) algorithm will be considered to account for missing values or other form of imputation method unless deletion of missing data is more ideal.

Conditions

Interventions

Eligibility

Outcomes