Study overview

This prospective observational study was approved by the Institutional Review Board of Partners Healthcare (now MassGeneralBrigham, Inc.) and all aspects of the study were performed in accordance with relevant guidelines and regulations. Patients were enrolled between November 19, 2018 and August 28, 2019 in the Cognitive Outcomes of Geriatric Surgery (COGS) Study, an ongoing investigation of contributors to and outcomes of postoperative delirium in patients aged 45–60 (middle-age) or ≥ 70 (older) years having elective spine surgery at Brigham and Women’s Hospital (BWH; Boston, MA). Details of COGS have been reported previously⁴² and recruitment and retention details are shown in a flow chart (Supplementary Fig. ¹). In brief, we included patients aged 45–60 or ≥ 70 years of age with an ASA physical status of I-III presenting for elective spine surgery and excluded those with a history of stroke or brain tumor; vision or hearing impairment that would impair ability to see pictures or read/hear instructions); limited use of the dominant hand (impaired drawing ability); or inability to speak, read, or understand English. Other than a prior diagnosis of dementia, there were no exclusion criteria for education or cognitive function. Of the 278 eligible patients approached, 69 declined to participate and 9 were ineligible, for a recruitment rate of 72% (N = 200). Of those, 7 did not have surgery (2 middle-aged, 5 old), 6 unenrolled (3 middle-aged, 3 old), and 105 had insufficient plasma in the biorepository at one or both timepoints. Finally, 6 patients were excluded due to missing data on necessary covariates, leaving 76 subjects (N = 34 middle-aged, 42 older) with preoperative and postoperative samples for the proteomic analysis.

After obtaining written informed consent, study staff collected baseline demographic information (age, sex, BMI, highest level of education) from the electronic medical record (EMR) preoperatively and administered the MiniCog, animal verbal fluency tests, and FRAIL scale^2,42. Delirium was identified by daily assessment with the 3-min diagnostic interview for CAM-defined delirium (3D-CAM)⁵² on postoperative days 1 to 3 and by chart review using published criteria^41,42. We used both methods because they are complimentary and well-established.

Delirium waxes and wanes so the 3D-CAM will miss delirium if it occurs at other times, whereas chart review reflects events over an entire day but may miss hypoactive delirium. The sensitivity and specificity of the combined method is superior to that of either method alone and maximizes identification of delirium⁴¹. As such, it is commonly used^13,42,43. The 3D-CM was administered in the patient’s room once per day in the morning prior to obtaining postoperative blood samples. Nearly all 3D-CAM assessments were performed by the same person (EJK), who was trained by a board-certified, clinically active geriatrician and was blinded to chart review information at the time of delirium screening. Patients were considered delirious if either the 3D-CAM or comprehensive patient chart review were positive⁴⁷. Surgical invasiveness, hospital length of stay, and 6-month mortality were collected by systematic chart review or examination of discharge diagnoses in the BWH Research Patient Data Registry. Invasiveness was categorized based on a scale of 1 to 4, where Tier 1 is least invasive (e.g. microdiscectomy) and 4 is most invasive (e.g. tumor, deformity, combined anterior and posterior procedures)^3,48. Data were collected and managed using Research Electronic Data Capture (REDCap), hosted at Partners Healthcare (Somerville, MA).