RESEARCHCognition and Capacity to Consent for Elective Surgery

Unrecognized cognitive disorders present ethical and logistical challenges for consenting older adults undergoing surgery. Twenty percent of adults aged 65 years or older have mild cognitive impairment or dementia, and the prevalence may be higher in perioperative patients.1-4 There is little literature on the epidemiology of preoperative incapacity, although diminished cognition correlates with impaired capacity for medical decision-making.5-7 As the population ages, it is imperative to recognize and manage patients with incapacity to consent for their upcoming surgeries. Studies suggest that physicians often overlook incapacity.8 In a cohort of 123 older adults presenting to a surgical ward, 111 (90.25%) had consented themselves for surgery, yet 33 (39.7%) had cognitive impairment, and 18 patients (16.2%) were unable to state the reason for admission to the hospital.9 Often, patients with profound impairment remain able to express a choice, without possessing an in-depth understanding of the attendant risks and benefits.6 Although performing formal capacity assessment on all patients in busy preoperative clinics is infeasible, a brief cognitive screening tool may allow for rapid identification of patients at highest risk of incapacity. The current pilot study had three aims: (1) determine feasibility of cognition and capacity assessment in a perioperative clinic; (2) describe the prevalence of incapacity in older adults presenting for surgery; and (3) examine the relationship between cognitive performance and capacity. This was a cross-sectional analysis performed at a single center, embedded in the Perioperative Optimization of Senior Health (POSH) quality improvement program. The POSH program is a collaborative care model between surgeons, geriatricians, and anesthesiologists, which has been described in detail elsewhere.10 This study received an exemption from the Duke Institutional Review Board. Patients aged 65 years or older and presenting for preoperative assessment with POSH in 2018 to 2019 were eligible for inclusion via convenience sampling. Exclusion criteria included: (1) non-English speaking; (2) hearing impairment that impeded communication; and (3) POSH appointment occurring less than 1 week before scheduled surgery. Participating surgical services included general, breast, gynecological, colorectal, hepatopancreaticobiliary, otolaryngology, cardiothoracic, orthopedics, and vascular. Cognition was assessed with the Montreal Cognitive Assessment (MoCA) and the Health and Safety subtest of the Independent Living Scale (ILS).11, 12 A subset of the MoCA items was used as an indicator of executive function, denoted as MoCA-EF.13 Patients with severe vision impairment were tested using the MoCA-BLIND.14 The ILS Health and Safety subtest primarily assesses judgment and executive function via awareness of potential hazards and hypothetical management of emergencies; it is scored from 0 to 40, with higher scores indicating better performance. Capacity to consent for surgery was assessed with the MacArthur Competence Assessment Tool for Treatment (MacCAT-T), a validated tool with excellent interrater reliability.15 The MacCAT-T evaluates patients' ability to (1) understand, (2) appreciate, (3) reason, and (4) express a choice, and generates scores for each of the four domains. There is no absolute cutoff determining incapacity; however, it provides a standardized approach for assessing capacity. A single assessor (K.E.Z.) performed all capacity assessments and was blinded to cognitive testing scores. If a participant was found to lack capacity, the participant, next of kin, POSH providers, and surgeon were notified. Statistical analysis was performed using RStudio (RStudio Inc). Fifty participants were enrolled in the study, and nine (18%) lacked capacity to consent for surgery. Median age was similar in the two groups (75 vs 76 years). Two patients who lacked capacity (22.2%) were female, compared with 27 (65.9%) of patients with capacity (P = .02). Patients who lacked capacity had a mean of 8.3 years of formal education (standard deviation (SD) = 5.7 years), compared with 14.9 years (SD = 3.1 years) for patients with capacity (P = .01). Figure 1 illustrates the receiver operator characteristic curves for MoCA, MoCA-EF, and ILS for predicting incapacity. The area under the curve for each test was 0.97, 0.88, and 0.79, respectively. At a cutpoint of 19 or less, the MoCA had 89% sensitivity and 93% specificity for predicting incapacity. At a cutpoint of 8 or less, the MoCA-EF had 88% sensitivity and 70% specificity for predicting incapacity. Deploying a brief cognitive screening test to older adults undergoing surgery may help identify those patients at highest risk of incapacity. All participants, including those with severe visual impairment, were able to complete the MoCA or MoCA-BLIND. At a cutpoint of 19, the MoCA had excellent sensitivity and specificity for predicting incapacity. The MoCA-EF also had excellent sensitivity and fair specificity. Screening with only the components of the MoCA-EF would potentially decrease the testing administration time. However, independently administering only the executive function components of the MoCA has not been validated, and is not possible for patients with severe visual impairment. Performing preoperative cognitive screening on adults older than 65 years aligns with recommendations for best practice by the American Geriatrics Society and American College of Surgeons.16 Patients scoring 19 or less on the MoCA merit further capacity evaluation. Identifying these high-risk patients allows surgeons to dedicate extra time to a complete capacity assessment, either with formalized tools appropriate for the clinical setting or through informal interviews to assess each domain of capacity.15 If patients are found to lack capacity, obtaining consent from the appropriate next of kin or healthcare power of attorney is essential before proceeding with surgery. Education was unbalanced between groups with and without capacity; however, the MacCAT-T emphasizes teach back and allows the interviewer to repeat or rephrase information appropriate to the patient's level of understanding. Patients with incapacity were much more likely to be male, but given the small sample size, the significance of this finding is unclear. This was a small study, performed at a single academic institution, and the POSH clinic is a specialized referral clinic, all of which may limit generalizability. In this pilot study, 18% of older adults presenting for elective surgery lacked capacity to consent for their upcoming procedure. Patients who scored 19 or less on the MoCA were at highest risk for incapacity. This was a small, single-center study; however, our data suggest the MoCA can be useful to identify older adults undergoing surgery who are at the highest risk of incapacity. Because the MoCA requires a fee for use, similar cognitive screening tools should also be examined for their ability to identify incapacity in older adults. The authors would like to acknowledge the efforts of the Perioperative Optimization of Senior Health clinical team, including Julie Little, NP, Mary “Betsy” Hixon, NP, Neema Sharda, MD, Serena Wong, DO, and Angeline Smith, LPN, for their assistance in facilitating this study. This project was funded by a Maddox Award, provided from the Duke Center for Aging, Duke University Medical Center. The authors have no relevant conflicts of interest to disclose. Kahli E. Zietlow participated in the study concept and design, acquisition of data, analysis and interpretation of data, and preparation of manuscript. Deborah Oyeyemi participated in the acquisition of data, analysis and interpretation of data, and preparation of manuscript. Sarah Cook participated in the study concept and design, analysis and interpretation of data, and preparation of manuscript. Margaret Hardy participated in the acquisition of data and preparation of manuscript. Shelley R. McDonald participated in the analysis and interpretation of data and preparation of manuscript. Sandhya Lagoo-Deenadayalan participated in the study concept and design and preparation of manuscript. Mitchell T. Heflin participated in the study concept and design, analysis and interpretation of data, and preparation of manuscript. Heather E. Whitson participated in the study concept and design, analysis and interpretation of data, and preparation of manuscript. None.