Customize your JAMA Network experience by selecting one or more topics from the list below.
Balentine CJ, Naik AD, Berger DH, Chen H, Anaya DA, Kennedy GD. Postacute Care After Major Abdominal Surgery in Elderly Patients: Intersection of Age, Functional Status, and Postoperative Complications. JAMA Surg. 2016;151(8):759–766. doi:10.1001/jamasurg.2016.0717
Advanced age is an important risk factor for discharge to postacute care (PAC) facilities including skilled nursing and rehabilitation. Factors modifying the age-related risk of discharge to PAC have not been adequately examined for surgical patients.
To evaluate how preoperative functional status and postoperative complications affect age-related risk of discharge to PAC facilities following major abdominal surgery.
Design, Setting, and Participants
Retrospective cohort study of 55 238 patients aged 18 years or older having colorectal, pancreas, or liver operations in 2011 and 2012 at hospitals participating in the National Surgical Quality Improvement Program. Age was classified as younger than 65 years, 65 to 74 years, 75 to 84 years, and 85 years or older. The study was conducted between July 1, 2014, and July 1, 2015.
Main Outcomes and Measures
The primary outcome was discharge to a PAC facility following surgery. The secondary outcome was type of PAC facility (skilled nursing, rehabilitation, or other facility).
Among 55 238 patients (mean [SD] age, 61  years; 49% male) having colorectal, pancreas, or liver operations, 5325 (10%) were discharged to PAC facilities after major abdominal surgery. Skilled nursing facilities were the most common type of PAC (63%), followed by rehabilitation hospitals (30%) and other facilities (7%). Older age was an important predictor of discharge to PAC facilities, but there were significant interaction effects with age and postoperative complications. Among functionally independent patients who avoided postoperative complications, rates of discharge to PAC increased from 1% in the group younger than 65 years to 30% in the group aged 85 years or older. For functionally independent patients with multiple complications, 13% of patients younger than 65 years were discharged to PAC facilities compared with 66% of those aged 85 years or older. After risk adjustment, the oldest patients were 27 times more likely to be discharged to PAC than the youngest group when there were no postoperative complications (odds ratio = 26.6; 95% CI, 21.6-32.7) and 11 times more likely after multiple complications (odds ratio = 11.4; 95% CI, 8.3-15.6). Among functionally dependent patients, the overall risk of discharge to PAC facilities was increased, but age was not as important a predictor for discharge to PAC.
Conclusions and Relevance
Older patients are frequently discharged to PAC facilities even when they are functionally independent and without postoperative complications. Helping older patients to return home after surgery and avoid placement in PAC facilities will require innovative programs that go beyond reducing complication rates and enhance postoperative recovery.
One of the defining demographic trends in the United States is an increase in the elderly population. There are currently more than 19 million Americans aged 65 to 74 years. There are also more than 18 million individuals older than 75 years, and this age group will account for 12% of the population by 2050.1,2 Older patients have unique difficulties with recovery from complex surgery because physiological reserve and the tolerance for adverse events decline with age. Consequently, it is more difficult for older patients to recover after surgery, regain independence, and return home after hospital discharge.3 Forty-five percent of those older than 65 years will have ongoing care needs after hospital discharge and require postacute care (PAC) services such as home health care, skilled nursing facilities, and inpatient rehabilitation.4,5
Being discharged to PAC facilities rather than returning home after surgery has important implications for patients and health systems. Discharge to a facility rather than home adds more than $10 000 to each episode of care, and Medicare spends $62 billion annually on PAC.6,7 From the patient’s perspective, discharge to a PAC facility rather than home is associated with up to a 4-fold increase in mortality and adds considerable stress for families and caregivers due to disruptions of normal routines.8,9 Consequently, the American College of Surgeons and The Joint Commission both recognize the ability to return home after surgery as a vital patient-centered measure of surgical quality.10,11
Although increasing age is associated with greater risk of discharge to PAC facilities, other factors also play an important role. Preoperative functional status is a key predictor of outcomes for older patients. Dependence in 1 or more activities of daily living (ADLs) increases mortality to 17% during hospitalization and up to 55% by 1 year after discharge.12 Poor preoperative function is also an important predictor of postoperative recovery and the need for PAC.13 Additionally, postoperative complications can derail recovery and lead to ongoing care needs that require formal support services through placement in PAC facilities.4,5,8 To effectively use PAC facilities after surgery, it is important to identify high-risk patients so they can be targeted for interventions to improve recovery prior to or during the inpatient hospital stay.1 Effective discharge planning requires early identification of patients who will need PAC services in order to provide adequate time for arranging resources prior to discharge.14 It is also important to help patients understand the potential implications of surgery including the risk of being unable to return home immediately after hospital discharge. Unfortunately, there is little information on how age, functional status, and postoperative complications interact to affect the risk of discharge to PAC after abdominal surgery. The purpose of this study is to evaluate how the age-related risk of discharge to PAC facilities after major abdominal surgery is affected by preoperative functional status and postoperative complications.
Question How is the age-related risk of discharge to postacute care affected by preoperative functional status and postoperative complications?
Findings In this cohort study of 55 238 patients having major abdominal surgery, the risk of discharge to postacute care facilities (skilled nursing, inpatient rehabilitation) rather than home varied significantly depending on preoperative functional status and postoperative complications.
Meaning To help older patients return home after surgery rather than being discharged to postacute care facilities, it will be important to enhance preoperative preparation and postoperative recovery programs and to reduce complication rates.
We conducted a retrospective cohort study using data from the 2011 and 2012 National Surgical Quality Improvement Program (NSQIP) Participant Use Data Files. The study was conducted between July 1, 2014, and July 1, 2015, under the auspices of a standing institutional review board protocol at the University of Wisconsin that covers use of all NSQIP data. We included patients aged 18 years or older having pancreatectomy, hepatectomy, or colorectal resection who were discharged home or to PAC facilities. We excluded anyone who was pregnant, had disseminated cancer, underwent another procedure within 30 days prior to admission, was transferred from a PAC facility, or was ventilator dependent before surgery. Informed consent was not required because the data set is publicly available and deidentified.
The primary outcome was discharge to any PAC facility, including skilled or unskilled nursing care, inpatient rehabilitation, or separate acute care. The secondary outcome was the type of PAC facility.
Age was divided into discrete groups: younger than 65 years, 65 to 74 years, 75 to 84 years, and 85 years or older. The 3 categories of functional status (independent, partially dependent, fully dependent) were collapsed into 2 categories (independent, partially or fully dependent) because the total number of dependent patients was relatively low. The NSQIP complication variables were summed to create a total complication count, and this was divided into 3 groups (0, 1, or ≥2 complications) because this breakdown had previously been shown to affect risk of discharge to PAC facilities.4 Because there were significant interaction effects between age, functional status, and number of postoperative complications, we stratified the analysis based on these factors.
Proportions were compared using χ2 test. Multivariable logistic regression models were constructed to estimate the relationship between age and odds of discharge to PAC facilities, adjusting for confounders. Differences were considered statistically significant if α < .05 without correction for multiple comparisons. Analysis was performed using SPSS version 23 statistical software (SPSS, Inc).
We identified 55 238 patients (mean [SD] age, 61  years; 49% male) who had colorectal, pancreas, or liver resection at NSQIP hospitals in 2011 and 2012. Among them, 5325 patients (10%) were discharged to PAC facilities. As expected, the prevalence of comorbidities was greater in older patients compared with younger patients (Table). Rates of discharge to PAC facilities increased steadily from 3% in patients younger than 65 years to 11% in those aged 65 to 74 years, 24% in those aged 75 to 84 years, and 42% in those aged 85 years or older. Overall postoperative complication rates also increased with age. Among patients younger than 65 years, 18% experienced 1 complication, while 10% had multiple complications. The overall complication rate in those aged 85 years or older was 43%, with 16% having multiple complications. Similarly, the percentage of patients who were partially or totally dependent in ADLs increased from 1% in the youngest group to 10% in the oldest.
Figure 1 illustrates the relationship between age and risk of discharge to PAC facilities. For patients who were functionally independent before surgery, the relationship between age and risk of discharge to PAC facilities was consistent (Figure 1A). Increasing age was associated with a stepwise increase in the risk of discharge to PAC facilities regardless of how many postoperative complications were encountered. Having more postoperative complications increased the absolute rate of discharge to PAC facilities but did not significantly change the pattern of steady increase in risk with older age.
A substantial number of functionally independent older patients were discharged to PAC facilities even when they avoided postoperative complications. In the absence of postoperative complications, rates of discharge to PAC increased from 1% in the group younger than 65 years to 30% in the group aged 85 years or older (P < .001). For independent patients having multiple complications, the risk of discharge to PAC facilities was even more significant: 13% of patients younger than 65 years were discharged to PAC facilities compared with 66% of those aged 85 years or older (P < .001).
Figure 1B demonstrates a very different relationship between age, complications, and risk of discharge to PAC facilities among patients who were either partially or totally functionally dependent in ADLs before surgery. The overall rate of discharge to PAC facilities was considerably increased compared with patients who were functionally independent. Even among the youngest patients with no postoperative complications, 18% were discharged to PAC facilities rather than returning home after surgery. More than half of patients in the groups aged 75 to 84 years (53%) and 85 years or older (51%) were discharged to PAC facilities even when they avoided postoperative complications. The rate of discharge to PAC facilities was 78% for the oldest patients who experienced multiple complications, and the rates ranged from 30% to 67% after only 1 complication. Additionally, the stepwise relationship between age and risk of discharge to PAC facilities largely disappeared when examining functionally dependent patients. For patients with multiple complications, there was no real difference in rates of discharge to facilities between patients in the 3 oldest age groups. The group aged 75 to 84 years and those aged 85 years or older also experienced essentially identical rates of discharge to PAC facilities in the absence of postoperative complications.
The most common discharge destination for patients with PAC needs was skilled nursing facilities, which made up 63% of all PAC discharges. Rehabilitation centers (30%) and other facilities (7%) were used for the remaining patients with PAC needs. Figure 2 shows the choice of PAC setting for functionally independent and dependent patients according to the number of postoperative complications. For fully independent patients with no complications (Figure 2A), skilled nursing facilities accounted for 54% of all PAC discharges in the youngest group and 73% in the oldest, while discharge to rehabilitation hospitals decreased from 31% to 25% of all PAC discharges, respectively. The youngest functionally independent patients with multiple complications (Figure 2E) were discharged to skilled nursing facilities in 54% of cases and rehabilitation in 36% compared with 72% and 23%, respectively, in the oldest group. Among patients with functional dependence who avoided postoperative complications (Figure 2B), the youngest patients used skilled nursing facilities for 58% of PAC cases vs 64% in the oldest group. The same population was discharged to rehabilitation 32% of the time for the youngest patients and 31% for the oldest. After multiple complications (Figure 2F), functionally dependent patients younger than 65 years were sent to skilled nursing facilities 53% of the time compared with 84% of those aged 85 years or older; rehabilitation for PAC needs was used for 33% of those younger than 65 years compared with 13% of those aged 85 years or older.
We calculated the odds of discharge to PAC facilities for the different age groups according to functional status and number of postoperative complications. Formal tests of interaction between age, functional status, and number of complications confirmed the need to examine these groups as distinct categories. Figure 3A and B illustrate how age and postoperative complications influenced the risk of discharge to facilities among patients who were functionally independent. In the absence of postoperative complications, the oldest patients were 35 times more likely to require PAC facilities than the youngest group (odds ratio [OR] = 35.2; 95% CI, 29.5-42.0) (Figure 3A). As the number of complications increased among functionally independent patients, the age effect remained significant but was attenuated. The oldest group was 16 times more likely to need PAC facilities at discharge compared with the youngest group after a single complication (OR = 15.6; 95% CI, 12.4-19.4) and 13 times more likely after multiple complications (OR = 13.4; 95% CI, 10.1-17.6). After adjusting for comorbidities and length of stay, the age effect persisted and the overall pattern of increasing risk with greater age remained similar (Figure 3B). Odds of discharge to PAC facilities in the oldest group compared with the youngest were increased 27-fold in the absence of postoperative complications (OR = 26.6; 95% CI, 21.6-32.7), 14-fold following 1 complication (OR = 13.6; 95% CI, 10.5-17.6), and 11-fold after multiple complications (OR = 11.4; 95% CI, 8.3-15.6).
Although there was a predictable and steady increase in the odds of discharge to PAC facilities with increasing age and number of complications for functionally independent patients, there was a distinctly different risk pattern for those with dependence in ADLs (Figure 3C and D). Patients younger than 65 years still had the lowest chance of discharge to PAC facilities. Individuals in older groups tended to have very similar odds of discharge to PAC facilities, even when accounting for postoperative complications. Figure 3C shows that the odds of discharge for the 3 oldest groups were generally indistinguishable regardless of whether they experienced 0, 1, or multiple complications. The oldest patients were 5 times more likely than the youngest patients to need PAC facilities in the absence of postoperative complications (OR = 4.8; 95% CI, 2.7-8.6), 5 times more likely after 1 complication (OR = 4.6; 95% CI, 2.4-8.9), and 4 times more likely after multiple complications (OR = 4.0; 95% CI, 1.8-9.1). The ORs were similar for the next oldest groups as well. After adjustment for demographic characteristics and comorbidity, the overall pattern of age-related effects was similar (Figure 3D).
Older individuals compose an increasing proportion of patients having surgery, so it is important to understand their unique needs during postoperative recovery.1,2 Traditional quality metrics such as 30-day morbidity and mortality fail to fully describe outcomes in older patients who also value the ability to maintain independence, maximize quality of life, and return home after surgery rather than spending long periods in PAC facilities.
We found that age is an important predictor of PAC needs, but the relationship is complex. Even older patients with excellent preoperative functional status who recover without complications may not be able to return home after surgery. Patients aged 85 years or older who were fully independent and had an uncomplicated postoperative course were discharged to skilled nursing or rehabilitation hospitals 30% of the time, and those aged 75 to 84 years used PAC facilities in 14% of cases. These findings suggest that helping older patients to return home after surgery is not just a matter of reducing postoperative complications but will require active measures to assist with physical and mental recovery. Additionally, more than half of patients with partial or total functional dependence prior to surgery will be discharged to PAC facilities regardless of how many complications they experience. For these patients, preoperative conversations with patients and their families need to clearly lay out this risk to set reasonable expectations for recovery. For older patients having elective surgery, there is an opportunity to modify risk factors for discharge to PAC prior to having surgery. At the same time, discharge planning should begin early with these patients to accommodate their complex care needs. This is especially true following emergency surgery. Because there is no opportunity to modify risk factors prior to the operation, it is important to quickly identify patients who are at risk for discharge to PAC. If we identify patients with PAC needs early rather than later during their hospital stay, then there is adequate time for families and discharge planners to collaborate and identify the appropriate facility to address recovery needs.
We also found that most surgical patients were discharged to skilled nursing facilities to address their PAC needs, while relatively fewer patients were discharged to rehabilitation hospitals. Both types of PAC can provide rehabilitation services, but rehabilitation hospitals offer more intensive therapy.15 Given our findings that many older patients without complications still required PAC services, it is reasonable to ask whether greater use of rehabilitation services might be of benefit. Our previous work demonstrated that the most common reasons for PAC referrals in older patients are deconditioning and a need for rehabilitation.3 While the reason for PAC referral in the current data set is not clear, deconditioning could certainly explain the need for PAC in the absence of discrete complications. To our knowledge, there is currently no evidence on the cost-effectiveness of rehabilitation vs skilled nursing for postoperative recovery after abdominal surgery. This could be a fruitful area for future inquiry.
Most research on PAC has focused on a limited number of diagnoses such as heart failure, lung disease, stroke, and hip fracture.16-18 Because surgical patients have different needs and complications than medical patients, there is a need to better explore the role of PAC after abdominal surgery.19 Sacks et al5 recently evaluated the overall frequency of PAC after colectomy, pancreatectomy, and open aneurysm repair for Medicare patients. They found that 45% of these patients needed PAC after discharge. They also demonstrated that postoperative complications and functional status were important indicators of the need for PAC, but the study did not appear to assess for interactions among these important variables. Our findings suggested a statistically significant interaction effect among age, functional status, and complications such that fitting a single model with all of these variables would give inaccurate estimates. We also took a different approach by paying less attention to the effects of individual complications and focusing more on the overall complication burden. Furthermore, their analysis was limited to patients aged 65 years and older, while we demonstrated that even younger patients can have significant PAC needs. Mohanty et al10 reported on the American College of Surgeons NSQIP Surgical Risk Calculator designed to estimate risk of discharge to facilities after surgery. Their model is based solely on preoperative factors and is designed to assist planning prior to surgery. Our data take into consideration events that occur during the hospital stay (postoperative complications) because adverse events clearly affect the need for PAC.
Our study has several limitations. First, NSQIP captures the most common surgical complications but may miss some procedure-specific complications relevant for pancreatic and liver surgery. It is possible that patients categorized as having no complications may have a complication not captured by NSQIP. However, there does appear to be a gradual increase in risk of discharge to facilities when comparing the groups with 0, 1, and multiple complications, which suggests that the 0-complication category is still reasonably accurate. The variable for functional status in NSQIP is also a crude indicator of ADL function. It does not distinguish between which ADLs are impaired, so we have only a general indication of disability. Additionally, need for PAC is influenced by factors not captured in this data set. Social support, access to different types of PAC, insurance incentives, socioeconomic status, and treatment patterns of local physicians can influence discharge destination. Cognitive impairment, also an important determinant of discharge destination, is not captured in our data set.20 Finally, frailty is a measure of function that has been correlated with adverse outcomes following surgery.21 Although this is not directly measured in NSQIP, a modified frailty index has been constructed from NSQIP data.21 Incorporation of this index did not affect our conclusions (data not shown), but this is likely due to clustering at the lower end of the frailty spectrum rather than an indication that frailty is unrelated to PAC use.
Our study demonstrates that older patients are more frequently discharged to PAC facilities to help with recovery following major abdominal operations. However, there is a complex relationship between age, functional status, and postoperative complications that affects the odds of discharge to PAC facilities. Further work is necessary to identify how the decision about discharge destination is made and whether more patients can safely be sent home after abdominal surgery.
Corresponding Author: Courtney J. Balentine, MD, MPH, Department of Surgery, University of Wisconsin, 600 Highland Ave, K4/729, Madison, WI 53792 (email@example.com).
Accepted for Publication: February 22, 2016.
Published Online: May 4, 2016. doi:10.1001/jamasurg.2016.0717.
Author Contributions: Dr Balentine had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Drs Anaya and Kennedy are co–senior authors.
Study concept and design: Balentine, Chen, Anaya, Kennedy.
Acquisition, analysis, or interpretation of data: Balentine, Naik, Berger.
Drafting of the manuscript: Balentine, Anaya.
Critical revision of the manuscript for important intellectual content: Balentine, Naik, Berger, Chen, Kennedy.
Statistical analysis: Balentine.
Administrative, technical, or material support: Berger.
Study supervision: Berger, Chen, Anaya, Kennedy.
Conflict of Interest Disclosures: None reported.
Funding/Support: This work was supported by a Career Development Award from Conquer Cancer Foundation, American Society of Clinical Oncology (Dr Anaya) and based on work at the VA Health Services Research and Development Center for Innovation in Quality, Effectiveness, and Safety, Michael E. Debakey VA Medical Center supported by grant CIN 13-413 from the Office of Research and Development, Veterans Health Administration, US Department of Veterans Affairs.
Role of the Funder/Sponsor: The funders 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 reflect the position or policy of the US Department of Veterans Affairs, Baylor College of Medicine, the American Society of Clinical Oncology, or the Conquer Cancer Foundation.
Create a personal account or sign in to: