Are functional, cognitive, and psychological measures that are grounded in geriatric assessment associated with 1-year mortality in older adults after major surgery?
In this cohort study, 17% of participants who underwent major surgery died within 1 year. Functional, cognitive, and psychological measures were significantly associated with mortality.
Specific measures, such as preoperative function, cognition, and psychological well-being, may need to be incorporated into the preoperative assessment to enhance surgical decision-making and patient counseling.
More older adults are undergoing major surgery despite the greater risk of postoperative mortality. Although measures, such as functional, cognitive, and psychological status, are known to be crucial components of health in older persons, they are not often used in assessing the risk of adverse postoperative outcomes in older adults.
To determine the association between measures of physical, cognitive, and psychological function and 1-year mortality in older adults after major surgery.
Design, Setting, and Participants
Retrospective analysis of a prospective cohort study of participants 66 years or older who were enrolled in the nationally representative Health and Retirement Study and underwent 1 of 3 types of major surgery.
Major surgery, including abdominal aortic aneurysm repair, coronary artery bypass graft, and colectomy.
Main Outcomes and Measures
Our outcome was mortality within 1 year of major surgery. Our primary associated factors included functional, cognitive, and psychological factors: dependence in activities of daily living (ADL), dependence in instrumental ADL, inability to walk several blocks, cognitive status, and presence of depression. We adjusted for other demographic and clinical predictors.
Of 1341 participants, the mean (SD) participant age was 76 (6) years, 737 (55%) were women, 99 (7%) underwent abdominal aortic aneurysm repair, 686 (51%) coronary artery bypass graft, and 556 (42%) colectomy; 223 (17%) died within 1 year of their operation. After adjusting for age, comorbidity burden, surgical type, sex, race/ethnicity, wealth, income, and education, the following measures were significantly associated with 1-year mortality: more than 1 ADL dependence (29% vs 13%; adjusted hazard ratio [aHR], 2.76; P = .001), more than 1 instrumental ADL dependence (21% vs 14%; aHR, 1.32; P = .05), the inability to walk several blocks (17% vs 11%; aHR, 1.64; P = .01), dementia (21% vs 12%; aHR, 1.91; P = .03), and depression (19% vs 12%; aHR, 1.72; P = .01). The risk of 1-year mortality increased within the increasing risk factors present (0 factors: 10.0%; 1 factor: 16.2%; 2 factors: 27.8%).
Conclusions and Relevance
In this older adult cohort, 223 participants (17%) who underwent major surgery died within 1 year and poor function, cognition, and psychological well-being were significantly associated with mortality. Measures in function, cognition, and psychological well-being need to be incorporated into the preoperative assessment to enhance surgical decision-making and patient counseling.
In the United States, more than 4 million operations are performed annually for patients 65 years and older.1 Older adults are more medically complex and are at higher risk of morbidity and mortality than younger adults.2 Currently, routine risk assessments emphasize medical conditions. However, health and well-being in older persons may be as much dictated by physical, cognitive, and psychological function as by medical conditions. These domains of function are not often part of routine risk assessment in surgical patients. Yet, a focus on medical conditions may not be adequate for the more medically complex and frail population. Conceptual models specific to geriatric care suggest that it is a combination of multiple domains of risk factors, such as physical, cognitive, and psychological function, that affects outcomes.
Despite high rates of older adults undergoing surgery, to our knowledge, the risk factors that may be especially relevant to outcomes in older adults who undergo a major surgery are not well studied.3-8 Specifically, a relevant outcome of long-term mortality beyond the traditional 30-day mortality better informs the patient’s surgical decision-making. For example, an asymptomatic frail, older adult may forgo a procedure, such as an abdominal aortic aneurysm repair (AAAR), if they have a high 1-year mortality risk. Improving our understanding of functional, cognitive, and psychological risk factors in this population, particularly their ability to predict risk beyond typical medical factors, is essential to providing patient-centered care.9-14 This understanding helps inform targets for preoperative risk assessment and optimization that may lead to improved patient outcomes. Additionally, incorporating these factors in risk adjustment will aid in appropriately comparing outcomes across clinicians who care for medically complex and frail patients.
The objective of this study was to understand functional, cognitive, and psychological risk factors that are associated with long-term mortality outcomes in older adults after major surgery. The goal is to inform the preoperative evaluation of the medically complex and frail population. This hypothesis was tested by using the Health and Retirement Study (HRS), a nationally representative cohort of older community-dwelling adults.
Data Sources and Patients
The HRS is a longitudinal study supported by the National Institute on Aging that measures changes in the health and economic circumstances of Americans as they age and is nationally representative of persons older than 50 years. Initiated in 1992, new participants have periodically been recruited to remain representative of the US population. From 1992 to 2014, HRS has recruited 12 waves of 18 000 to 23 000 participants each and continues to recruit. The HRS sample is based on a multistage area probability design involving geographical stratification and the clustering and oversampling of certain demographic groups.15 HRS interviews are conducted by phone or face to face (overall response rate, >80%) every 2 years. If an individual is unable to complete an interview because of physical or cognitive impairment, the interview is conducted with a proxy respondent, generally a family member.
We identified HRS participants 66 years and older who underwent a major surgery case by linking the HRS survey to Medicare claims. We used a prior classification that defined major surgery as either an AAAR, coronary artery bypass graft (CABG), or colectomy. This combination of operations has been used in prior studies of high-risk surgery because they are common, high risk and reliably coded in claims data and together represent a diverse range of operations.16,17 High risk was defined as having a 30-day mortality rate of at least 1%.1 These procedures were identified by International Classification of Diseases, Ninth Revision (ICD-9) codes (AAAR: 3834, 3844, 3925, and 3864; CABG: 3610, 3611, 3612, 3613, 3614, 3615, 3616, 3617, 3619, 362, 363, 3631, 3632, 3633, 3634, and 3639; colectomy: 4571, 4572, 4573, 4574, 4575, 4576, 4579, 4581, 4582, 4583, 1731, 1732, 1733, 1734, 1735, 1736, 1739, 4503, 4526, 4541, and 4549). Of 28 013 HRS participants 66 years or older between 1992 and 2014, 24 647 (88%) agreed to have their HRS surveys linked to the Medicare claims. We identified 2291 participants who underwent a major surgery. Because we used Medicare claims to identify comorbidities before the major surgical event, we excluded participants who were not enrolled in Medicare fee for service 1 year before the surgery. Of the remaining 1829 participants, 488 (27%) who had no HRS interview within 2.5 years before the major surgery were excluded. The resulting cohort included 1341 HRS participants who underwent a major surgery. The institutional review board at the University of California, San Francisco, reviewed and approved this study. Participants or a proxy provided written informed consent for the HRS.
Data Collection and Measurement
The HRS interview data were used to characterize the sample in terms of self-reported age, sex, race or ethnicity (eg, non-Hispanic white, non-Hispanic black, Hispanic, and other), education, wealth, net worth, income, marital status, nursing home residence, and presence of a geriatric syndrome (eg, hearing or vision difficulty, incontinence). We determined a history of dementia and a Charlson comorbidity score for each participant using Medicare claims.
The primary outcome was postoperative mortality at 1 year following surgery. We did additional analyses that examined 6-month and 2-year mortality. Mortality was determined using National Death Index–linked data.
Variables were chosen with the recognition that geriatric surgical outcomes may be because of the complex interplay of patients’ medical, functional, and psychological risk. Variables were derived from the HRS participant interview preceding the surgery. Three functional measures were included. Activities of daily living (ADL) included difficulty and a need for help with bathing, dressing, eating, using the toilet, getting in and out of bed, and walking across the room. Instrumental ADL (IADL) included difficulty and a need for help with preparing meals, financing, using a phone, shopping, and taking medication. For ADL and IADL, participants were first asked if they had difficulty doing the activity. If they had difficulty, they were then asked if they were able to complete the activity without help. For ADL and IADL function, participants were classified into 1 of the following hierarchal categories: able to complete all the activities with or without difficulty in 1 or more activities (independent); a need for help in 1 activity (dependent in 1 ADL); or a need for help in more than 1 activity (dependent in 2 or more ADLs). Finally, we considered whether participants reported having difficulty walking several blocks. We included a psychological measure to assess the presence of depression. Depression was measured based on the 8-item Center for Epidemiologic Studies–Depression (CES-D) interview. Depression was defined by a CES-D score of 3 or more.18 This (CES-D) interview was not incorporated into the HRS questionnaire until 1993; therefore, participants who underwent surgery before this time were not included in the depression analysis.
Cognitive function was measured using the validated HRS cognitive scale, which was derived from Mini-Mental State Examination (MMSE) and telephone interview of cognitive status. When the participant was not able to be interviewed, a validated algorithm using surrogate observations of cognitive capacity was used. Based on the cognitive scale or surrogate observations, participants were classified as cognitively normal, cognitively impaired, or with dementia using the Langa-Weir Cognitive algorithm developed by Langa et al19 and validated with HRS participants.20 If the Langa-Weir cognitive score was missing, the Medicare ICD-9 code for dementia (290.XX) was used.
Our analytic approach sought to determine whether measures of physical, cognitive, and psychological function are associated with mortality after accounting for comorbidity burden, demographic characteristic (age, race/ethnicity, sex, net worth, and years of education), and surgery type. We used Cox regression models to determine associations between geriatric factors and mortality through 2 years of follow-up and to estimate hazard ratios and 95% confidence intervals.
We adjusted for patient age, sex, race/ethnicity, education, income, wealth, Charlson comorbidity score, and surgery type. Because surgical techniques may have changed over the years, we performed a sensitivity analysis adjusting for the year the surgery was performed. The findings were not significant and are therefore not presented in this article. To illustrate the clinical effect of each of these risk factors, we used these models to find the association between 6-month, 1-year, and 2-year mortality for patients with and without each risk factor after adjustment for the other variables in the model. To estimate the association of accumulating geriatric risk factors, we then categorized the patients by the number of geriatric risk factors they had and determined the percentage who died at 1 year as a function of the number of risk factors.
All analyses were weighted to account for the differential probability of participant selection and the complex survey design of the HRS. Statistical analyses were performed SAS, version 9.3 (SAS Institute) and statistical significance was set at P < .05.
The characteristics of the 1341 older adults in the HRS who underwent major surgery are presented in Table 11,19 (mean [SD] age, 76  years; 737 female [55%]; 1134 white [85%]). Most were community-dwelling older adults (1316 [98%]), with 847 (63%) married or partnered and 805 (60%) who had social support other than their spouse. Eight hundred eighty-four (66%) had more than 2 comorbidities. Most were independent or had difficulty but did not need help in ADLs and IADLs (1238 [92%] and 1170 [90%], respectively). Seventy-five (6%) had dementia and 271 (23%) had cognitive impairment without dementia. Three hundred six (25%) had depression prior to surgery.
Mortality After Major Surgery
Of the 1341 participants who underwent high-risk surgery, 225 (17%) died within a year of surgery. The hazard ratios of the probability of death and the calculated 6-month, 1-year, and 2-year rates of death are presented in Table 2.1,19 A graph of the mortality rates in each group can be found in the eFigure in the Supplement. After adjusting for age, sex, race/ethnicity, education, income, wealth, and surgery type, the following factors were associated with 1-year mortality: multimorbidity, dependence in 2 or more ADL or IADLs, not being able to walk several blocks, cognitive function, and depression. The Figure shows the 1-year mortality rate increases markedly as the number of geriatric risk factors increase, increasing from 10% in those with no risk factors to 28% in those with 2 or more risk factors.
In this nationally representative study of older persons who underwent 1 of 3 major surgical operations, 17% died within a year of the operation. Measures that are crucial to the well-being of older persons, including ADL impairment, IADL impairment, mobility impairment, dementia, and depression, were strongly associated with 1-year mortality even after adjusting for multimorbidity and demographic factors. The accumulation of geriatric risk factors was associated with higher mortality rates. Our findings support the notion that preoperative assessment in older persons needs to consider not just the disease burden of the patient but also these domains of physical and cognitive functioning and psychological well-being.
To our knowledge, this is the first study to evaluate the association between multiple risk factors grounded in the geriatric assessment with long-term postoperative outcomes using a nationally representative longitudinal cohort of older adults who have undergone major surgery. In our previous work in the nursing home population, we showed that functional and cognitive impairment were strongly associated with 1-year mortality and functional decline in women undergoing surgery for breast cancer.21 Other studies, generally single-centered, have found that preoperative physical and cognitive function are associated with short-term mortality (ie, 30-day, 6-month).22-24 One national study found that preoperative cognition and function was associated with 30-day mortality in older surgical adults.25 This increasing body of literature supports the importance of functional and cognitive assessments in the preoperative setting, specifically for older adults undergoing major surgery.
While psychological risk factors are important to older adults and are their own domain in the comprehensive geriatric assessment, few have assessed the association of psychological factors with mortality in older adults undergoing major surgery. We identified that 25% had depression before their surgery and depression was significantly associated with 1-year mortality. In a study of CABG surgery, 30% to 40% of patients were affected by depression and the risk associated with mortality after CABG surgery increased independently of medical factors.26
Strengths and Limitations
The key strength of our study was our ability to examine multiple nontraditionally evaluated risk factors in a nationally representative cohort of older persons undergoing major surgery. Because of the variable timing of HRS interviews, a limitation of our study is that measures were assessed at variable points before surgery depending on the timing of the HRS interview. However, increased time between measuring the geriatric risk factors and surgery would blunt the association between these risk factors and surgical mortality, assuming that the measures are worse the more proximal to the surgery. Thus, our study likely understates the importance of these risk factors. There may be several limitations to using the HRS and the development of our cohort. While the HRS cohort is developed to be nationally representative of older adults in the United States, it may not be representative of the elderly US population undergoing surgery. Analyses using a subsample of the HRS are likely to be somewhat less representative than the overall sample. Also, as is common in analyses uses Medicare claims, HRS participants who underwent surgery while enrolled in a Medicare Advantage program were excluded. Therefore, the representativeness of our surgical sample may be less than that of the surgical population in the United States. In regard to the development of the surgical cohort, we selectively chose 3 major operations, AAAR, CABG, and colectomy, but did not include other types of operations, such as the total hip replacement, which is a very common operation yet relatively low-risk operation. Because of this, for example, the population we described may not be representative of those that undergo the total hip replacement. Lastly, because our data lacked the reason of death, we are unable to determine the main cause of the mortality and are therefore unable to discuss the potential association of the main cause of death with the measures of interest. However, 242 of the HRS participants who underwent a colectomy had colorectal cancer (43.5%) and this may help explain the higher mortality in this cohort.
Increasingly, the literature supports that factors, such as cognition and function, are associated with surgical outcomes. Our research findings highlight the need to include the assessment of cognition, function, and psychological factors in the preoperative setting for older adults undergoing major surgery. These findings matter for several reasons: they highlight the importance of evaluating nonmedical factors in conjunction with the traditional factors and support the need for a paradigm shift in preoperative medical care to a more holistic approach in which cognition, function, and psychological factors are considered. Risk factors found to be significant likely increase the baseline risk of mortality in those with and without surgery; however, identifying these risk factors in the preoperative setting will likely help identify those who have less ability to withstand the stress of surgery. This information will help inform those who may have a limited life expectancy following surgery to consider whether the stress and potentially lengthy recovery period may outweigh any benefits when surgery may be elective. These findings justify including nontraditional risk factors in risk models and risk assessments and may be used to identify potential modifiable risk factors with the goal of improving postoperative outcomes.
Caring for the medically complex and frail older patients needs to incorporate the evaluation of functional, cognitive, and psychological factors in the evaluation of outcomes and risk assessments. A study of patients in an orthopedic spine clinic revealed that 70% of their cohort had preoperative cognitive impairment and those with cognitive impairment had worse postoperative delirium and postoperative complications and discharge institutionalization.27 This highlights that cognitive impairment is prevalent and associated with poor outcomes, yet it is not usually assessed in the preoperative setting in surgery clinics. Future research should examine how to incorporate geriatric factors in the preoperative setting using the increasing literature we have to frame decision-making within the surgical setting and how to use this information of poor postoperative outcomes in older adults with geriatric risk factors in anticipatory guidance. Additionally, studying whether these geriatric factors are modifiable and potential targets for intervention in the preoperative setting will be essential to changing care and improving outcomes.
Functional, cognitive, and psychological risk factors are associated with postoperative mortality. These risk factors, grounded in the geriatric assessment, add to surgical decision-making and anticipatory guidance and may inform future potential interventions. Incorporating these factors into the preoperative assessment is the first next step to improving care for older adults undergoing major surgery.
Accepted for Publication: January 16, 2020.
Corresponding Author: Victoria L. Tang, MD, MAS, Division of Geriatrics, Department of Medicine, University of California, San Francisco, 4150 Clement St, 181(G), San Francisco, CA 94121 (firstname.lastname@example.org).
Published Online: March 11, 2020. doi:10.1001/jamasurg.2020.0091
Author Contributions: Dr Tang had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Tang, Finlayson, Covinsky, Boscardin.
Acquisition, analysis, or interpretation of data: Tang, Jing, Boscardin, Ngo, Silvestrini, Covinsky, Finalyson.
Drafting of the manuscript: Tang.
Critical revision of the manuscript for important intellectual content: Tang, Jing, Boscardin, Finlayson, Covinsky.
Statistical analysis: Jing, Boscardin.
Obtained funding: Tang.
Administrative, technical, or material support: Ngo, Silvestrini.
Supervision: Finlayson, Covinsky, Boscardin.
Conflict of Interest Disclosures: Dr Finlayson reports grant support from the National Institutes of Health and being a founding shareholder of Ooney Inc. No other disclosures were reported.
Funding/Support: This work was supported by the National Institute on Aging Grants for Early Medical/Surgical Specialists’ Transition to Aging Research (R03AG056342), the National Institute on Aging Mentored Career Development Award (KL2TR001870), and the National Institute on Aging of the National Institutes of Health and the University of California, San Francisco Claude D. Pepper Older Americans Independence Center (P30AG044281).
Role of the Funder/Sponsor: The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Disclaimer: The views expressed in this article are those of the authors and do not necessarily represent those of the National Institute on Aging or National Institutes of Health.
Meeting Presentation: These findings were presented at the Annual Meetings of the Society of General Internal Medicine; April 11, 2018; Denver, Colorado; American Geriatrics Society; May 2, 2019; Portland, Oregon; and American College of Surgeons; October 27, 2019; San Francisco, California.
et al. Predictors of long-term survival after coronary artery bypass grafting surgery: results from the Society of Thoracic Surgeons Adult Cardiac Surgery Database (the ASCERT study). Circulation
. 2012;125(12):1491-1500. doi:10.1161/CIRCULATIONAHA.111.066902PubMedGoogle ScholarCrossref
et al; ACE trialists. A randomized controlled trial of endovascular aneurysm repair versus open surgery for abdominal aortic aneurysms in low- to moderate-risk patients. J Vasc Surg
. 2011;53(5):1167-1173.e1. doi:10.1016/j.jvs.2010.10.124PubMedGoogle ScholarCrossref
et al. Risk associated with complications and mortality after urgent surgery vs elective and emergency surgery: implications for defining “quality” and reporting outcomes for urgent surgery. JAMA Surg
. 2017;152(8):768-774. doi:10.1001/jamasurg.2017.0918PubMedGoogle ScholarCrossref
et al; PREDIMED Study Investigators. Impact of psychosocial factors on cardiovascular morbimortality: a prospective cohort study. BMC Cardiovasc Disord
. 2014;14:135. doi:10.1186/1471-2261-14-135PubMedGoogle ScholarCrossref
et al. Psychosocial risk and protective factors and incident mild cognitive impairment and dementia in community dwelling elderly: findings from the Singapore Longitudinal Ageing Study. J Alzheimers Dis
. 2017;57(2):603-611. doi:10.3233/JAD-160862PubMedGoogle ScholarCrossref
JS. Costs and consequences of early hospital discharge after major inpatient surgery in older adults. JAMA Surg
. 2017;152(5):e170123. doi:10.1001/jamasurg.2017.0123PubMedGoogle Scholar
TN. The significance of preoperative impaired sensorium on surgical outcomes in nonemergent general surgical operations. JAMA Surg
. 2015;150(1):30-36. doi:10.1001/jamasurg.2014.863PubMedGoogle ScholarCrossref
FL, de Farias
TP, Albuquerque Sousa
LCS. Predictors of postoperative complications, prolonged length of hospital stay, and short-term mortality in elderly patients with malignant head and neck neoplasm. ORL J Otorhinolaryngol Relat Spec
. 2014;76(3):153-164. doi:10.1159/000363189PubMedGoogle ScholarCrossref
et al. The prevalence of undiagnosed pre-surgical cognitive impairment and its post-surgical clinical impact in elderly patients undergoing surgery for adult spinal deformity. J Spine Surg
. 2017;3(3):358-363. doi:10.21037/jss.2017.07.01PubMedGoogle ScholarCrossref