Study design
FINDERI is a prospective, single-center, observational pilot trial. The primary hypothesis is to validate a delirium screening questionnaire for POD. The primary hypothesis is to validate a delirium screening questionnaire for POD. Secondary, the study aims to identify specific pre- and perioperative risk factors and biomarkers for POD and cognitive decline in patients undergoing cardiac surgery, and to investigate if cardiac surgery accelerates a cognitive decline in the presence of Aβ pathology. Ethical approval for the study was obtained from the IRB of the University of Göttingen Medical Center (#20/11/20). The investigators registered this study in German Clinical Trials Register (DRKS) (DRKS00025095) on April, 19th 2021
Study setting and participants
In total, 500 patients aged over 50 years old undergoing cardiac surgery will be recruited at the Department of Cardiovascular and Thoracic Surgery of the University Medical Center Göttingen, Germany. Inclusion and exclusion criteria are summarized in Table 1. Potentially eligible participants are identified by preoperative schedules. Informed consent will be collected after provided detailed study information and prior to baseline assessment and biomarker sampling. Initially, the preoperative screening will be provided via medical history, screening questionnaire, and laboratory measurement for identification of risk factors for delirium, cognitive decline, and dementia (t0). Furthermore, delirium screening will be assessed over the first five postoperative days (t1) and biomarkers will be measured on the 7th postoperative day (t2) to address the impact on neurodegenerative and neuroinflammatory markers after cardiac surgery. We will provide a follow-up visit of patients with a verified diagnosis of POD 3 to 5 months after cardiac surgery (t3), follow-up of all patients on phone 12 months after surgery (t4), and follow-up of patients with a verified diagnosis of POD 18 months after surgery (t5) (Fig. 1, Table 2).
Preoperative screening (t0)
Upon admission, a preoperative delirium screening is carried out using a questionnaire (Supplement 1) developed in the context of the multicenter cluster-randomized PAWEL study[36] on the reduction of delirium risk and postoperative cognitive dysfunction after elective procedures in older adults. The preoperative screening questionnaire includes the following assessment: A) Geriatric check (mobility, statutory level of independency, cognition, psychological symptoms, and previous hospital stay); B) Short 6-item cognitive screening; C) General information including results from A) and B) as well as age > 80, laboratory measurement such as increased creatine, CRP, hemoglobin, decreased protein, American Society of Anesthesiologists (ASA) score ≥ 3 [37]. The patient's current medications are also noted; however, particular attention is paid to the number of medications and medications with a deliriogenic risk. Furthermore, pre-surgical diagnosis of depression, stroke, dementia, and previous hospital stays within the last year will be assessed. In addition, it is checked whether a care level exists and whether the patient is a nursing home resident. The patient is asked about previous delirium and a recent increase in number of falls. Another component of the survey is the self-reported subjective memory impairment (SMI), alcohol use, and smoking status. Finally, the patient's handgrip strength will be measured by Jamar® Hydraulic Hand Dynamometer.
Furthermore, the following variables will be evaluated by trained assessors for baseline assessments (Supplement 2). Basic sociodemographic information will be collected, including marital status, number of children, immigrant background, educational level, occupation, living arrangement, and care level. Moreover, an examination of previous cardiac and noncardiac comorbidities will be provided. Frailty will be assessed using the 7-point Clinical Frailty Scale of the Canadian Study of Health and Aging (CSHA Frailty Scale, 1-very fit up to 7-severely frail) [38]. In addition, the outcome-oriented nursing assessment instrument AcuteCare 1.1 (ePA-AC) [39], whereby the software generates a delirium risk score and a self-care score, will be assessed from the MEONA software. Factors biasing the adequate elicitation of the ePA-AC (Supplement 3), such as the understaffing of nurses, will be documented.
The pre-surgical cognitive status will be measured using neuropsychological screening, including Montreal Cognitive Assessment (MoCA), Trail Making Test A, and Trail Making Test B (TMTA, TMTB). MoCA is a 30-point brief cognitive screening with the following subscales: short-term memory recall task (5 points), visuospatial abilities (5 points), executive functions (4 points), attention, concentration, and working memory (5 points), language (5 points), orientation to time and place (6 points). The MoCA has high sensitivity (0.90) and specificity (0.87) to detect patients with mild cognitive impairment [40]. TMT is a widely used neuropsychological instrument to assess the speed of cognitive processing, visuomotor tracking, divided attention, and cognitive flexibility [34, 41–43]. The test consists of two parts (A and B). In part A, patients are instructed to draw lines connecting consecutively numbered circles from 1 to 25 in ascending order. In part B, the circles include both numbers (1–13) and letters (A-L), and the patients draw lines to connect the circles in an ascending pattern, but with the added task of alternating between the numbers and letters. Total time will be recorded in seconds. In elderly volunteers, part B had a specificity of 0.89, and sensitivity of 0.63 for cognitive dysfunction, and 0.72 for dementia [34, 41–43].
Finally, blood samples will be measured before the surgery to address the hypotheses and predictive values of serum biomarkers for delirium, cognitive decline, and early-onset dementia. This examination section will take approximately 60 minutes.
Delirium screening (t1) and postoperative laboratory measurements (t2)
All study patients will be assessed twice a day using the Confusion Assessment Method (CAM) on the intensive care unit (CAM-ICU) or I-CAM on the intermediate care unit. CAM is a four features diagnostic algorithm for delirium screening at the bedside with a rating of acute onset or symptom fluctuation, inattention, disorganized thinking, and altered level of consciousness. I-CAM has a high sensitivity of 0.77 in a cohort of geriatric patients with a high prevalence of dementia and a specificity of 0.96-1.00 with inter-rater reliability of Cohen´s kappa 0.95. CAM-ICU is an adaptation of I-CAM for critically ill patients on or off the ventilator. The first step in assessing delirium in ICU is the Richmond Agitation-Sedation Scale (RASS), followed by the four features of the I-CAM with an extra assessment of pictures in feature 2 if patients are not able to complete letters. CAM-ICU shows a high sensitivity of 0.95-1.00 and specificity of 0.89–0.93 with inter-rater reliability of Crohn´s kappa from 0.88 to 1.0 [44–47]. Since POD symptoms often occur at night, chart reviews[48] (Supplement 4) will be assessed for the first five postoperative days. This section of the investigation will take approximately 10–15 minutes per survey. Blood will be sampled again on the 7th postoperative day (t2). Finally, a surgery protocol using the data from cardiac surgery and anesthesiology will be collected using the following data: type of cardiac surgery, time of surgery (cut-to-suture-time), aortic cross-clamp time, intraoperative complications, and use/time of the heart-lung machine.
Neurocognitive assessments (t3)
Patients suffering from POD within 5 postoperative days will be asked to continue the study for additional follow-up visits (3 to 5 months after cardiac surgery) for in-depth memory assessments (t3). They will undergo a standardized neuropsychological investigation. Neuropsychological assessments are performed to evaluate the patients’ neurocognitive stage to define different stages of cognitive decline, namely mild cognitive impairment or dementia. A diagnosis of dementia (DSM-5: major neurocognitive disorder) or mild cognitive impairment (DSM-5: mild neurocognitive disorder) is made according to the 5th edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) [49]. A neuropsychological battery will be applied to determine cognitive impairment or early dementia (Table 3): MoCA, Subtests of the Wechsler Adult Intelligence Scale IV (WAIS-IV: Block Design, Digit Span forward and backward, Coding), subtests of the Wechsler Memory Scale IV (WMS-IV: Logical Memory I and II, Visual Reproduction I and II), the Rey-Osterrieth Complex Figure Test (RCFT), the CERAD (Consortium to Establish a Registry for Alzheimer's Disease)-Plus test battery (including Trail Making Test A and B and letter fluency), clock drawing test and a test of ideomotor apraxia (imitation of finger gestures) by Goldenberg. The estimated time for full neuropsychological testing will be approximately 75 minutes.
Test results will be discussed with a physician at the memory clinic at the Department of Psychiatry and Psychotherapy of the University of Göttingen Medical Centre. If indicated, additional diagnostic procedures will be offered to determine a clinical diagnosis. Neuropsychiatric symptoms will be measured using established and standardized questionnaires and external rating scales, depending on the clinical presentation. The following instruments (Table 4) can be used: Movement Disorder Society-Unified Parkinson's Disease Rating Scale (MDS-UPDRS), REM Sleep Behavior Disorder Screening Questionnaire (RBDSQ), Neuropsychiatric Inventory (NPI), and Scale for the Assessment of Positive Symptoms (SAPS). In addition, another blood sample is taken.
Follow up (T4 after 12 months and T5 after 18 months)
One year after the cardiac surgery, all patients (with and without delirium after cardiac surgery) will be contacted via phone to evaluate any new onset of neurological disorders, deterioration of cognitive decline using MoCA or physical autonomy, and current medications. Furthermore, previous COVID-19 infection and vaccination status will be assessed. The telephone interview will take approximately 30 minutes.
After 18 months, another standard neuropsychological examination will be provided in hospital for patients with a verified diagnosis of POD after cardiac surgery. This visit is comparable with the t3 visit, including standard neuropsychological battery for assessment of cognitive impairment or early dementia (Table 3), and neuropsychiatric symptoms questionnaire depending on the clinical presentation (Table 4).
Blood sampling and biomarker measurements
Blood sampling and biomarker analysis are some of the main goals and hypotheses of this study. Up to 50 ml of blood per study visit (t0, t2, t3, t5) will be sampled. Blood collection (serum, plasma, RNA out of PAXgene tubes), processing, and storage will be performed according to local established SOPs. Serum and plasma will be briefly kept at 4°c until processing. Serum will be stored for 45 minutes for coagulation. Serum and EDTA-plasma will be centrifugated for 10 minutes at 2,000 x g and stored as 500µl aliquots at -80°c. The cellular pellet from the EDTA-tube will be reconstituted in PBS and stored at − 80°c for DNA extraction. PAXgene tube will be stored for 120 minutes at room temperature, followed by 24 hours at -20°c and long-term storage at -80°c until use. The investigation is initially designed as a prospective single sample survey and longitudinal data collection. With the help of various protein-analytical or other measurement methods, new biomarkers for POD, cognitive decline, and dementia will be identified and validated. In addition to proteomic or metabolomic markers, genomic and epigenetic markers, such as the ApoE genotype, will be examined. The test persons are therefore explicitly advised that a genetic examination of the donated material may be carried out. To protect the participants, no feedback is given about individually achieved markers, in particular predictive markers. To test our hypothesis that pre-surgical Alzheimer’s pathology is a risk factor to develop POD followed by cognitive decline, we will measure plasma Aβ using our recently developed two-step immunoassay [65, 66], p-tau181 [67, 68], NfL [69, 70], and GFAP[26] using SIMOA technology [71, 72]. Samples will be handled according to our recently published SOP suggestions to minimize pre-analytical effects on plasma Aβ [73]. As POD and cognitive decline are often considered to be associated with a concomitant increase in inflammatory cytokines, we want to identify markers that correlate with delirium and neurodegenerative markers. Therefore, we will also analyze a cytokine panel, including IL-1ra-, IL-6, IL-8, IL-10, and tumor necrosis factor (TNF) alpha [23, 74–76]. Furthermore, to overall understand the impact of specific cardiovascular alteration and to test the hypothesis that changes in cardiovascular markers are associated with delirium, we analyze the cardiovascular marker growth differentiation factor 15 (GDF-15) [77].
Study sample
In order to plan a subsequent confirmatory study, it is necessary to estimate the necessary parameters with sufficient accuracy. The sample size planning is based specifically on estimating the area under the curve (AUC) of the delirium short screening for the occurrence of delirium with a 95% confidence interval so that the 95% confidence interval has a width of approximately 0.05 points. It is assumed that the prevalence of POD is 50% and that there is a true AUC of 0.7. In 416 patients analyzed, the 95% confidence interval extends approximately 0.025 points from the estimate. In order to compensate for possible dropouts (assumed dropout rate approx. 20%), 500 patients are therefore recruited. The calculation was done in nQuery 8.
Statistical analysis
In order to determine the prognostic properties of constant endpoints and questionnaire scores to predict the development of POD, as well as postoperative cognitive decline, Alzheimer's dementia, and prognostic quality measures are calculated. Specifically, an ROC curve of the observations with the associated AUC with optimal cut-off point (simultaneous maximization of sensitivity and specificity, as well as according to Youden) will be performed. Furthermore, positive and negative predictive values for the optimal cut-off point and the observed incidence risk factors for the development of POD, cognitive decline, and Alzheimer's or Lewy body dementia will be investigated using logistic regression and effect sizes are reported as hazard ratios. As far as possible, all key figures are given with a 95% confidence interval. In addition to the univariate consideration of the individual parameters, a multivariate consideration of predictive factors is also planned. For this purpose, several machine learning methods (supervised learning) are trained on a training data set (70% of the data collected) and their accuracy is compared with a test data set (30% of the data collected). The division into test and training data sets is random.
Data management
All outcomes are entered into a Good Clinical Practice (GCP) compliant database (secuTrial®), configured for this trial. The configuration includes univariate checks for plausibility, such as range checks. Data are regularly reviewed for completeness by qualified personnel and locked after review. A blinded data review (without knowledge on the development of delirium) to assess data quality is performed prior to database lock. After database lock, the data set is archived for at least 10 years on a digital medium within the trial master file. An anonymized copy of the data set is provided alongside the publication to ensure the reproducibility of results. The investigators follow the Findability, Accessibility, Interoperability, and Reuse (FAIR) Guiding Principles for scientific data management and stewardship.