Sage Journals: Discover world-class research

Abstract

Background

Ostomy reversal is a common surgical procedure; however, it is not without associated risks. Patient selection for this elective procedure is therefore critically important. Elderly patients represent a growing population and a substantial proportion of patients that present for evaluation after ostomy creation due to the most common etiologies. This study aims to assess the impact of frailty on the outcomes of ostomy reversal among older adults.

Methods

Patients ≥65 years who underwent ostomy reversal from 2015 to 2019 were identified in the NSQIP database. Frailty was calculated using the 5-item Modified Frailty Index (MFI). Multivariate regression was performed to evaluate the association of frailty with post-operative 30-day mortality, 30-day serious complications, discharge to a facility, and 30-day readmission.

Results

A total of 13,053 patients were included, of which 18.7% were frail (MFI ≥ 2). Patients who underwent colostomy reversal had higher rates of serious complications (P < .0001) and discharge to facility (P < .0001) compared to other reversals. In multivariate analysis, frailty was associated with increased odds of serious complications (OR 1.52, 95% CI 1.31-1.77), discharge to facility (OR 2.14, 95% CI 1.79-2.57), and readmission (OR 1.23, 95% CI 1.04-1.46), but not mortality. Frail patients had predicted probabilities 1.4 times higher for serious complications and 1.7-2.2 times greater for discharge to facility than non-frail patients.

Conclusions

Among older adults undergoing elective ostomy reversal, frailty is independently associated with increased odds of 30-day serious complications, discharge to facility, and 30-day readmission. As a potentially modifiable risk factor, identification of frailty offers the opportunity for shared decision-making and prehabilitation.

Keywords

colorectal general surgery geriatrics

Introduction

In an aging US population, patient selection for elective surgery has become increasingly important since advanced age is a risk factor for increased post-operative morbidity and mortality. Older adults have worse outcomes than younger patients across surgical specialties, which is partially due to burden of comorbidities. Therefore, the decision to offer elective surgery to older adults can be complex, and the risks and benefits of the procedure must be taken into account. There is growing awareness, however, that age alone is not always a reliable predictor of post-operative outcomes. The concept of frailty has seen increasing attention in the literature, as this factor incorporates health parameters beyond chronological age such as physiologic reserve, social support, and psychological status.^1,2 Multiple frailty metrics have been developed with significant variation in the complexity of different instruments.^3–6 The 5-item Modified Frailty Index (MFI), which accounts for comorbidities and functional status, has been shown to predict outcomes in both general^7–9 and colorectal surgery.^10–12

Ostomy creation is common in general and colorectal surgery and can be performed in both the elective and emergent setting. Common indications for ostomy creation include diverticular disease and colorectal cancer, which are each seen with increasing frequency in older populations.^13,14 While ostomy creation is often necessary at the time of the original procedure to reduce perioperative complications, the presence of an ostomy can be associated with decreased quality of life which may prompt discussions about ostomy reversal.¹⁵ Elective ostomy reversal is highly desired by most patients;¹⁶ however, it is not without associated morbidity, and therefore, careful evaluation of the risk/benefit ratio is needed in vulnerable older adults.¹⁷ Prior studies examining the outcomes of ostomy closure have included a general population,^18–21 but few have focused on older adults.^22,23 Few prior studies have specifically evaluated the role of frailty in predicting outcomes of ostomy closure in older adults.²⁴

Therefore, the aim of this study is to assess the impact of frailty on post-operative outcomes of older adults undergoing elective ostomy reversal. Specifically, the study identifies the impact of frailty, as measured by the 5-item MFI, on the perioperative outcomes of 30-day mortality, 30-day serious complications, discharge to skilled nursing facility, and 30-day readmission. We hypothesize that frailty is associated with increased negative outcomes among older adults undergoing elective ostomy reversal.

Methods

Data Source

This was a cohort study of patients undergoing elective ostomy reversal using the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) database. The NSQIP registry includes prospectively collected surgical patient data from over 700 hospitals including demographics, perioperative variables, and 30-day outcomes. This study was deemed exempt by the Institutional Review Board of the University of California Los Angeles.

Patient Selection

Patients aged ≥65 years who underwent ostomy reversal between 2015 and 2019 were identified by principal Current Procedural Terminology (CPT) code for ostomy reversal (see Table 1). Patients were excluded if their operation was designated as emergent. International Classification of Diseases (ICD) 9 and 10 codes were used to classify the type of ostomy as ileostomy, colostomy (see Table 1), or unknown (all other ICD codes, nonspecific to type of ostomy) (see Supplementary Data). Therefore, patients undergoing ileostomy reversal were defined by a CPT code for stoma reversal in combination with an ICD code mentioning the presence of an ileostomy. Patients undergoing colostomy reversal were similarly defined by a CPT code for stoma reversal in combination with an ICD code mentioning the presence of a colostomy. A laparoscopic approach was defined by the CPT code 44227. Patients with concurrent procedures were included. A small proportion (8.1%) of the study population had concurrent CPT codes, of which the majority were cystourethroscopy and/or ureteral stent implantation (65.2%). Therefore, we did not adjust for concurrent procedures in the analysis.

Table 1.

CPT Codes for Ostomy Reversal Used as Inclusion Criteria for Study Population and ICD Codes for Ileostomy and Colostomy Subgroups.

CPT codes	Description
44620	Closure of enterostomy, large or small intestine
44625	Closure of enterostomy, large or small intestine; with resection and anastomosis (other than colorectal)
44626	Closure of enterostomy, large or small intestine; with resection and colorectal anastomosis (eg, closure of Hartmann-type procedure)
44227	Laparoscopy, surgical, closure of enterostomy, large or small intestine, with resection and anastomosis
ICD codes—ileostomy
V44.2^a	Ileostomy status
V55.2^a	Attention to ileostomy
Z43.2^b	Encounter for attention to ileostomy
Z93.2^b	Ileostomy status
ICD codes—colostomy
V44.3^a	Colostomy status
V55.3^a	Attention to colostomy
569.62^a	Mechanical complication of colostomy
K94.00^b	Colostomy complication, unspecified
K94.03^b	Colostomy malfunction
K94.09^b	Other complications of colostomy
Z43.3^b	Encounter for attention to colostomy
Z93.3^b	Colostomy status

CPT = Current Procedural Terminology; ICD = International Classification of Diseases.

^aICD-9.

^bICD-10.

Primary Regressor of Interest

The 5-item MFI was the primary regressor of interest. The 5-item MFI is calculated using 5 variables: diagnosis of congestive heart failure, diagnosis of diabetes, diagnosis of chronic obstructive pulmonary disease, diagnosis of hypertension requiring medication, and dependent functional status (either “partially” or “completely” dependent).⁵ The presence of each item is given 1 point, for a total score of 0-5. Frailty is defined as a score of ≥2.

Other Predictor Variables

Patient characteristics of age, sex, race/ethnicity, body mass index (BMI), and active tobacco use (tobacco user within 1 year of surgery) were included as predictor variables. Age was categorized into 3 groups: 64-74 years, 75-84 years, and 85+ years. Body mass index was categorized into 2 groups based on the cutoff for defining obesity: BMI <30 kg/m² and BMI ≥30 kg/m². Additionally, preoperative laboratory values of serum creatinine (mg/dL) and albumin (g/L) levels were also included.

Outcomes

The outcomes of interest were 30-day mortality, 30-day serious complications, discharge to a facility, and 30-day readmission. The definition of serious complications was based on the ACS NSQIP surgical risk calculator, which defines 30-day serious post-operative complications as cardiac arrest, myocardial infarction, renal insufficiency, acute renal failure, reintubation, pneumonia, venous thromboembolism, wound dehiscence, deep incisional or organ space surgical site infections (SSIs), systemic sepsis, urinary tract infection, or reoperation.²⁵ A subanalysis of serious complications was performed to evaluate differences in individual complications by ostomy procedure.

Statistical Analysis

STATA version 16 (Statacorp, College Station, TX) was used to perform all analyses. Univariate analysis was performed using a chi-square test for categorical variables and Mann-Whitney U tests for continuous variables which were all non-normally distributed. Univariate logistic regression was performed for the 3 outcomes of interest. Variables with statistical significance in the univariate regression were included in the respective multivariate models. Patients with missing values for variables in the regression analysis were not included in the final regression. Predicted probabilities were calculated with coefficients from the multivariate models for frail and non-frail adults at each age category using marginal analysis.

Results

Bivariate Analysis

A total of 13,053 older adults underwent elective ostomy reversal, of which 53.3% were female with a median age of 71. Overall, 18.7% of the total cohort was frail. Patient characteristics, preoperative laboratory values, and primary outcomes for the entire cohort in addition to stratification by non-frail and frail are displayed in Table 2. Frailty did not differ by type of ostomy reversal. Frail patients were older (P < .001), more often male (P < .0001), more likely to be of a racial or ethnic minority group (P < .0001), more likely obese (P < .0001), and more likely active smokers (P = .001) compared to non-frail patients. Patients who underwent colostomy reversal were older (P < .0001), more often male (P < .0001), predominantly non-Hispanic white (P < .0001), more likely obese (P = .0001), and more likely active smokers (P = .0001) compared to those who underwent ileostomy reversal.

Table 2.

Patient Characteristics and Outcomes for the Total Cohort Including Stratification by Type of Ostomy.

Variable	Total population (N = 13,053)	Non-frail (N = 10,617)	Frail (N = 2436)	P-value
Patient characteristics
Age: N (%)				<.0001
65-74	8741 (67.0%)	7197 (67.8%)	1544 (63.4%)
75-84	3761 (28.8%)	2976 (28.0%)	785 (32.2%)
>85	551 (4.2%)	444 (4.2%)	107 (4.4%)
Gender: N (%)				<.0001
Female	6953 (53.3%)	5748 (54.1%)	1205 (49.5%)
Male	6100 (46.7%)	4869 (45.9%)	1231 (50.5%)
Race/ethnicity: N (%)				<.0001
Non-Hispanic white	9405 (72.1%)	7777 (73.3%)	1628 (66.8%)
Non-Hispanic black	799 (6.1%)	565 (5.3%)	234 (9.6%)
Non-Hispanic Asian	326 (2.5%)	253 (2.4%)	73 (3.0%)
Hispanic	692 (5.3%)	510 (4.8%)	182 (7.5%)
Unknown	1831 (14.0%)	1512 (14.2%)	319 (13.1%)
BMI: N (%)
<30 kg/m²	9601 (73.5%)	8074 (76.1%)	1527 (62.7%)	<.0001
≥30 kg/m²	3452 (26.5%)	2543 (23.9%)	909 (37.3%)
Smoker: N (%)				.001
No	11445 (87.7%)	9356 (88.1%)	2089 (85.8%)
Yes	1608 (12.3%)	1261 (11.9%)	347 (14.2%)
Serum albumin (g/L) (median, IQR)	3.9 (3.6-4.2)	3.9 (3.6-4.2)	3.9 (3.6-4.2)	<.0001
Serum creatinine (mg/dL) (median, IQR)	.95 (.8-1.2)	.95(.8-1.2)	.96 (.8-1.2)	<.0001
Operation type: N (%)				.082
Ileostomy	5822 (44.6%)	4771 (44.9%)	1051 (43.1%)
Colostomy	3819 (36.4%)	3819 (35.9%)	935 (38.4%)
Unknown	2477 (18.9%)	2027 (19.1%)	450 (18.5%)
Open surgery: N (%)
No	1472 (11.3%)	1219 (11.5%)	253 (10.4%)	.123
Yes	11581 (88.7%)	9398 (88.5%)	2183 (89.6%)
Outcome measures
30-day mortality: N (%)				.038
No	12925 (99.0%)	10522 (99.1%)	2403 (98.7%)
Yes	128 (1.0%)	95 (.9%)	33 (1.4%)
30-day serious complication: N (%)				<.0001
No	11401 (87.3%)	9409 (88.6%)	1992 (81.8%
Yes	1652 (12.7%)	1208 (11.4%)	444 (18.2%)
Reoperation	618 (4.7%)	451 (4.3%)	167 (6.9%)	<.0001
Organ space SSI	510 (3.9%)	399 (3.8%)	111 (4.6%)	.067
Pneumonia	289 (2.2%)	176 (1.7%)	113 (4.6%)	<.0001
Urinary tract infection	251 (1.9%)	195 (1.8%)	56 (2.3%)	.134
Reintubation	171 (1.3%)	117 (1.1%)	54 (2.2%)	<.0001
Wound infection	254 (1.9%)	169 (1.6%)	85 (3.5%)	<.0001
Dehiscence	131 (1.0%)	87 (.8%)	44 (1.8%)	<.0001
Deep venous thrombosis	101 (.8%)	78 (.7%)	23 (.9%)	.287
Myocardial infarction	86 (.7%)	58 (.6%)	28 (1.2%)	<.0001
Renal insufficiency	76 (.6%)	52 (.5%)	24 (.9%)	.004
Renal failure	53 (.4%)	34 (.3%)	19 (.8%)	.001
Cardiac arrest	58 (.4%)	41 (.4%)	17 (.7%)	.037
Pulmonary embolism	49 (.4%)	36 (.3%)	13 (.5%)	.157
Cerebrovascular accident	39 (.3%)	26 (.2%)	13 (.5%)	.019
Discharge destination: N (%)				<.0001
Home	11897 (92.4%)	9842 (93.8%)	2055 (86.0%)
Facility	984 (7.6%)	650 (6.2%)	334 (13.9%)
30-day readmission				<.0001
Yes	11709 (89.7%)	9588 (90.3%)	2121 (87.1%)
No	1344 (10.3%)	1029 (9.7%)	315 (12.9%)

BMI = body mass index; IQR = interquartile range; MFI = Modified Frailty Index; SSI = surgical site infection. Bold signifies a statistically significant p-value.

In analysis of outcomes, there was no difference in 30-day mortality by ostomy procedure. Patients who underwent colostomy reversal did have higher rates of both serious complications (P < .0001) and discharge to facility (P < .0001) compared to those undergoing ileostomy reversal. Subanalysis revealed that reoperation, organ space SSI, and pneumonia were the most frequent serious post-operative complications (see Table 2), of which reoperation (P < .0001) and pneumonia (P < .0001) were significantly more likely to occur in frail patients. However, the 3 most common complications were significantly less likely to occur after ileostomy closure when compared to colostomy reversal or unknown operations. Rates of clostridium difficile were highest among ileostomy reversal (3.5%), compared to colostomy (2.0%) and unknown reversals (2.3%, P < .0001).

Univariate Regression Analysis

Results of univariate logistic regression for each of the 3 outcome measures are displayed in Table 3. Frailty was associated with increased odds of all 4 negative outcomes: 30-day mortality (OR 1.52, 95% CI 1.02-2.27), serious complications (OR 1.74, 95% CI 1.54-1.96), discharge to facility (OR 2.50, 95% CI 2.17-2.87), and 30-day readmission (OR 1.38, 95% CI 1.21-1.58). Of all predictors of 30-day serious complications, patient frailty had the largest effect size. Age groups 75-84 and ≥85 had increased odds of worse outcomes, with age ≥85 having the largest effect size. Obesity and smoking were also identified as potential risk factors associated with worse outcomes. With regard to procedure, colostomy reversal and unknown ostomy reversals had greater odds of serious complications and being discharged to a facility compared to ileostomy reversals.

Table 3.

Univariate Logistic Regression of Predictor Variables With the Outcomes of 30-Day Mortality, 30-Day Serious Complication, Discharge to Facility, and 30-Day Readmissions.

Patient characteristics	30-day mortality		30-day serious complication		Discharge to facility		30-day readmission
Patient characteristics	Univariate OR (95% CI)	P-value	Univariate OR (95% CI)	P-value	Univariate OR (95% CI)	P-value	Univariate OR (95% CI)	P-value
Frailty (MFI >2)	1.52 (1.02-2.27)	.039	1.74 (1.54-1.96)	<.0001	2.50 (2.17-2.87)	<.0001	1.38 (1.21-1.58)	<.001
Age
65-74	Reference 1.0		Reference 1.0		Reference 1.0		Reference 1.0
75-84	1.68 (1.15-2.46)	.008	1.23 (1.10-1.28)	<.0001	3.11 (2.71-3.58)	<.0001	.986 (.84-1.09)	.485
>85	4.58 (2.69-7.78)	<.0001	1.36 (1.07-1.73)	.012	4.88 (3.84-6.20)	<.0001	1.01 (.76-1.34)	.976
Female sex	.73 (.51-1.03)	.071	.94 (.85-1.05)	.290	1.45 (1.27-1.66)	<.0001	1.04 (.93-1.16)	.509
Ethnicity
Non-Hispanic white ethnicity	Reference 1.0		Reference 1.0		Reference 1.0		Reference 1.0
Non-Hispanic black ethnicity	1.05 (.51-2.17)	.902	.94 (.76-1.18)	.614	1.47 (1.16-1.85)	.001	.93 (.73-1.18)	.692
Non-Hispanic Asian ethnicity	1.29 (.47-3.52)	.625	.57 (.38-.87)	.008	.70 (.44-1.13)	.142	.73 (.49-1.10)	.136
Hispanic ethnicity	1.37 (.68-2.72)	.378	.87 (.68-1.11)	.249	.61 (.43-.85)	.005	.95 (.74-1.23)	.714
Unknown	1.03 (.63-1.69)	.908	1.18 (1.02-1.36)	.027	.51 (.41-.65)	<.0001	.90 (.76-1.07)	.225
BMI ≥30	.78 (.51-1.18)	.240	1.23 (1.10-1.38)	<.0001	2.46 (2.14-2.83)	<.0001	.99 (.87-1.13)	.816
Smoker	1.24 (.76-2.03)	.383	1.28 (1.11-1.46)	.001	.81 (.66-1.01)	.057	1.08 (.91-1.28)	.361
Laboratory values
Albumin, serum	.39 (.29-.52)	<.0001	.71 (.63-.78)	<.0001	.53 (.47-.60)	<.0001	.82 (.73-.92)	.001
Creatinine, serum	1.31 (1.17-1.47)	<.0001	1.11 (1.04-1.19)	<.0001	1.10 (1.01-1.19)	.021	1.08 (1.00-1.16)	.036
Ostomy type
Ileostomy	Reference 1.0		Reference 1.0		Reference 1.0		Reference 1.0
Colostomy	1.22 (.83-1.82)	.309	1.44 (1.28-1.62)	<.0001	1.77 (1.52-2.05)	<.0001	1.02 (.90-1.16)	.727
Unknown	1.27 (.82-1.97)	.285	1.58 (1.37-1.81)	<.0001	1.83 (1.53-2.18)	<.0001	.96 (.82-1.12)	.622
Open surgery	1.69 (.86-3.33)	.131	1.08 (.91-1.27)	.392	1.16 (.93-1.43)	.188	1.06 (.88-1.27)	.550

BMI = body mass index; CI = confidence interval; MFI = Modified Frailty Index; OR = odds radio. Bold signifies a statistically significant p-value.

Multivariate Regression Analysis

Results of multivariate logistic regression for each of the 4 outcomes measures are displayed in Table 4. Frailty was associated with increased odds of serious complications (OR 1.52, 95% CI 1.31-1.77), discharge to facility (OR 2.14, 95% CI 1.79-2.57), and readmission (OR 1.23, 95% CI 1.04-1.46), but not with mortality. Only low albumin level had higher odds of negative outcomes in all 4 models. Including length of stay (LOS) in the regression model revealed that the risk of serious complications increased by 4% (OR 1.04, P < .001) and the risk of discharge to facility increased by 8% (OR 1.08, P < .0001) for each additional day of hospital admission. Open surgery was associated with a higher risk of discharge to facility (OR 1.35, P = .033) compared to laparoscopic surgery, but the approach was not associated with the other outcomes. Performance of the regression models were assessed using area under the curve (AUC). The AUC for mortality, serious complications, discharge to facility, and readmission were .72, .62, .74, and .55, respectively.

Table 4.

Multivariate Logistic Regression for Primary Outcomes of Predictor Variables With the Outcomes of 30-Day Mortality, 30-Day Serious Complication, and Discharge to Facility.

Patient characteristics	30-day mortality		30-day serious complication		Discharge to facility		30-day readmission
Patient characteristics	Adjusted OR (95% CI)	P-value	Adjusted OR (95% CI)	P-value	Adjusted OR (95% CI)	P-value	Adjusted OR (95% CI)	P-value
Frailty (MFI >2)	1.21 (.74-1.97)	.457	1.52 (1.31-1.77)	<.0001	2.14 (1.79-2.57)	<.0001	1.23 (1.04-1.46)	.017
Age
65-74	Reference 1.0		Reference 1.0		Reference 1.0		Reference 1.0
75-84	1.51 (.95-2.41)	.082	1.18 (1.03-1.36)	.019	2.89 (2.43-3.45)	<.0001	.92 (.79-1.08)	.321
>85	3.75 (1.87-7.51)	<.0001	1.15 (.83-1.58)	.403	4.66 (3.41-6.38)	<.0001	.77 (.52-1.13)	.182
Female sex	.90 (.58-1.38)	.628	.96 (.84-1.09)	.514	1.49 (1.25-1.77)	<.0001	1.06 (.92-1.22)	.423
Ethnicity
Non-Hispanic white ethnicity	Reference 1.0		Reference 1.0		Reference 1.0		Reference 1.0
Non-Hispanic black ethnicity	.68 (.28-1.67)	.401	.75 (.58-.98)	.035	1.24 (.93-1.65)	.144	.78 (.58-1.04)	.093
Non-Hispanic Asian ethnicity	.44 (.06-3.23)	.423	.65 (.41-1.03)	.067	.84 (.49-1.45)	.529	.62 (.38-1.02)	.061
Hispanic ethnicity	1.61 (.76-3.41)	.211	.85 (.64-1.13)	.254	.58 (.38-.87)	.009	1.07 (.81-1.42)	.641
Unknown	.79 (.38-1.66)	.526	1.08 (.89-1.33)	.433	.53 (.38-.74)	<.0001	.92 (.73-1.16)	.480
BMI ≥30	.87 (.52-1.46)	.607	1.17 (1.01-1.35)	.033	1.38 (1.14-1.65)	.001	.96 (.82-1.13)	.664
Smoker	1.08 (.54-2.14)	.826	1.38 (1.15-1.66)	.001	1.06 (.80-1.39)	.695	1.07 (.87-1.32)	.525
Laboratory values
Albumin, serum	.40 (.29-.54)	<.0001	.72 (.64-.80)	<.0001	.56 (.49-.65)	<.0001	.82 (.73-.93)	.001
Creatinine, serum	1.35 (1.18-1.54)	<.0001	1.11 (1.03-1.20)	.005	1.09 (.99-1.21)	.077	1.05 (.96-1.15)	.253
Ostomy type
Ileostomy	Reference 1.0		Reference 1.0		Reference 1.0		Reference 1.0
Colostomy	1.49 (.91-2.46)	.113	1.64 (1.41-1.90)	<.0001	2.10 (1.72-2.56)	<.0001	1.08 (.92-1.27)	.341
Unknown	1.30 (.75-2.25)	.350	1.58 (1.33-1.86)	<.0001	2.05 (1.65-2.54)	<.0001	.94 (.78-1.13)	.512
Open surgery	1.45 (.65-3.21)	.362	1.23 (1.00-1.53)	.054	1.35 (1.02-1.78)	.033	1.09 (.86-1.37)	.469

BMI = body mass index; CI = confidence interval; MFI = Modified Frailty Index; OR = odds radio. Bold signifies a statistically significant p-value.

Predicted probabilities of outcomes stratified by age, frailty, and ostomy type are shown in Table 5. There were no significant differences in predicted probability by frailty of 30-day mortality for any age group. Frail patients were predicted to have serious complication rates about 1.4 times higher than non-frail patients for all age groups and ostomy procedures. For example, among patients who underwent ileostomy reversal, rates of serious complications were 40% higher for frail patients compared to non-frail patients, regardless of age group. Frail patients who underwent colostomy reversal also had 40% higher rates of serious complications compared with non-frail patients undergoing colostomy reversal. Frail patients were also found to have predicted probabilities 1.7-2.2 times greater than non-frail patients for discharge to facility across groups by age and ostomy procedure. The largest difference in discharge to facility was seen among patients 65-74 years old. Frail patients were predicted to have readmission rates significantly higher than non-frail patients.

Table 5.

Predicted Probabilities of Outcome Pressures by Age, Frailty Status, and Stoma Procedure.

Patient group	Predicted probability of 30-day mortality (%)	Predicted probability serious complication (%)	Predicted probability of discharge to facility (%)
Ileostomy
Non-frail 65-74 years	0.7	9.4	2.9
Frail 65-74 years	0.8	13.6	6.0
Non-frail 75-84 years	1.0	10.9	7.8
Frail 75-84 years	1.2	15.7	15.2
Non-frail 85+ years	2.5	10.6	12.0
Frail 85+ years	3.0	15.3	22.2
Colostomy
Non-frail 65-74 years	1.0	14.1	5.6
Frail 65-74 years	1.2	19.9	11.2
Non-frail 75-84 years	1.4	16.2	14.3
Frail 75-84 years	1.7	22.7	26.0
Non-frail 85+ years	3.4	15.8	21.1
Frail 85+ years	4.1	22.2	35.7
Unknown
Non-frail 65-74 years	0.9	13.9	5.7
Frail 65-74 years	1.0	19.7	11.3
Non-frail 75-84 years	1.3	16.0	14.4
Frail 75-84 years	1.6	22.4	26.1
Non-frail 85+ years	3.1	15.6	21.1
Frail 85+ years	3.7	22.0	35.8

Discussion

The current study seeks to provide information to support patient selection for an elective ostomy reversal among older adults. In this large, national database study of older adults undergoing ostomy reversal, frailty was independently associated with increased odds of adverse post-operative outcomes including 1.5 times the rate of serious complications, 2.2 times the rate of discharge to a skilled nursing facility, and 1.2 times the rate of readmission within 30 days. Age was also predictive of worse outcomes, with patients aged 85 years and older being 3.7 times more likely to die within 30 days after surgery and over 4.5 times more likely to be discharged to a facility compared to patients of ages 65-74 years. In addition, patient risk factors of obesity and tobacco use also predicted serious short-term complications. Finally, the type of ostomy reversal also impacts outcomes with colostomy reversal (which is most commonly reversal of an end colostomy) independently associated with increased odds of serious complications and discharges to facility when compared to ileostomy reversal. Frailty was associated with higher predicted probability of serious complications and discharge to facility among patients who underwent both ileostomy and colostomy reversal.

While age as a deciding factor is appealing in its simplicity, the literature is clear that the calculus is more complex. Frailty scoring has emerged as a valuable predictor since it incorporates both functional and physiologic domains, which may provide better guidance. The elective setting, such as an outpatient visit for consideration of ostomy closure, allows for in-depth discussions and shared decision-making based on risk factors such as frailty, the oldest old (age 85+), obesity, and active tobacco use as shown in the current study. In addition, there is opportunity for risk modification with preoperative optimization of comorbidities as well as prehabilitation to potentially improve exercise tolerance, nutrition, and smoking cessation. While ostomy reversal is often desirable from a quality of life standpoint, it is important to consider the risks of this elective surgery and make sure the risk/benefit ratio is concordant with the patient’s goals of care.

Our findings are consistent with prior studies showing the negative impact of frailty on post-operative outcomes.^7–12,24 Two recent systematic reviews have supported the finding that frailty is associated with worse outcomes in colorectal surgery. Maturana et al reviewed 17 studies and concluded that frail patients undergoing elective or emergent colorectal cancer surgery have more post-operative complications, and in 5 of 7 included studies found higher mortality rates among frail patients.¹² In a review by Fagard et al, frailty was again associated with more post-operative complications after primarily elective colorectal surgery, although there was no association with mortality.¹¹ Further, Al-Khamis et al used the ACS-NSQIP database to find that short-term outcomes after both elective and non-elective colorectal surgery were worse among frail patients, including higher morbidity, mortality, and discharge to facility.¹⁰ Our study findings are supported by these large investigations on the impact of frailty in colorectal surgery. A recent NSQIP analysis by Kooragayala et al found frailty (also defined by the 5-item MFI) predicted worse outcomes among patients undergoing Hartmann’s reversal.²⁴ Interestingly, frailty was not a predictor of mortality in our elective surgery study population, and it appears that in prior literature, studies including non-elective surgery found mortality differences by frailty score. This finding likely reflects selection bias in the current cohort of patients who were selected to undergo elective ostomy reversal and likely felt to have a low risk of perioperative mortality.

The current study also examined the role of chronological age as a risk factor. Prior literature shows that in the majority of cases, age is not an independent risk factor for morbidity following ostomy reversal.^{18,20,26–28} For example, several studies showed that age was not associated with increased morbidity after ileostomy^18,26 or Hartmann’s reversal.^20,28 Similarly, Bosshardt et al found that even age over 70 years was not predictive of complications following ostomy closure.²³ In our study population, however, older age was significantly associated with some negative outcomes. Patients aged 85 years and older had higher risk of mortality, those aged 75-84 were at an increased risk of serious complications, and those aged 75 and older had higher rates of discharge to a facility. Therefore, while in our analysis older age may be predictive of worse outcomes, these findings support the use of more nuanced risk factors such as frailty since age is a non-modifiable risk factor.

New guidelines from the American Society of Colon and Rectal Surgeons (ASCRS) support the use of frailty over chronological age in perioperative care.²⁹ These new guidelines recommend preoperative screening for frailty and consideration of prehabilitation based on patient-specific factors that may improve outcomes. Prehabilitation can target and improve individual risk factors in multiple domains including activity level and exercise endurance, nutrition and weight loss, and psychosocial dynamics. The new ASCRS guidelines also provide a strong recommendation for the integration of prehabilitation into surgical pathways for frail patients when possible.²⁹ While the MFI score for a given patient may not change after prehabilitation, screening for frailty using MFI identifies at-risk patients who would benefit from optimization with the potential to reduce perioperative risk.

This study does have several limitations. First, the use of a large, retrospective database is subject to the inherent limitations of the available data. A significant proportion of the patients could not be categorized by type of ostomy procedure (ie, ileostomy vs colostomy takedown) or by underlying disease process (ie, malignant vs benign) due to non-specific coding. Second, this study is subject to selection bias as only patients who ultimately underwent ostomy closure were included. Therefore, these selected patients are likely among the most fit for elective surgery, which is supported by the overall low mortality rates seen in our data. Furthermore, the effect sizes determined by this study are likely underestimated for the same reason. In addition, despite in the growing use of preoperative frailty evaluation, there is no consensus on the optimal instrument for frailty assessment. Therefore, although the 5-item MFI has been shown to predict outcomes in general and colorectal surgery, it has been also shown to underperform in its predictive value when compared to other frailty measures.³⁰ However, it was nevertheless correlated with more in-depth metrics such as the Risk Analysis Index, suggesting that MFI remains a valuable marker of frailty. Finally, the outcomes studied here are limited to short-term 30-day outcomes, while the long-term outcomes such as quality of life after major surgery and functional recovery after surgery are also important to consider.

This study showed that among older adults undergoing elective ostomy reversal, frailty predicts increased odds of both 30-day serious complications and discharge to a facility independent of other variables such as age. These results add to the mounting evidence for the use of frailty as well as chronological age for preoperative assessment of surgical risk. As a potentially modifiable risk factor, identification of frailty is important for shared decision-making regarding the risk/benefit ratio of surgery as well as the opportunity for prehabilitation in this vulnerable patient population. Future studies should evaluate the association of frailty with longer-term outcomes such as quality of life or functional recovery after surgery.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Dr. Salzman was supported by the VA office of Academic Affiliations through the National Clinician Scholars Program.

ORCID iDs

Caitlyn Braschi

Garrett A. Salzman

Data Availability Statement

The given this article are: The American College of Surgeons National Surgical Quality Improvement Program and the hospitals participating in the ACS NSQIP are the source of the data used herein; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors.

References

Lin

Watts

Peel

Hubbard

. Frailty and post-operative outcomes in older surgical patients: A systematic review. BMC Geriatr. 2016;16(1). doi:10.1186/S12877-016-0329-8

Xie

Larson

Gonzalez

Pressler

Lustig

Arslanian-Engoren

. Components and indicators of frailty measures: A literature review. J Frailty Aging. 2017;6(2):76-82. doi:10.14283/JFA.2017.11

Shah

Borrebach

Hodges

, et al. Validation of the risk analysis index for evaluating frailty in ambulatory patients. J Am Geriatr Soc. 2020;68(8):1818-1824. doi:10.1111/JGS.16453

Callahan

Clark

Edwards

, et al. Automated frailty screening at-scale for pre-operative risk stratification using the electronic frailty index. J Am Geriatr Soc. 2021;69(5):1357-1362. doi:10.1111/JGS.17027

Subramaniam

Aalberg

Soriano

Divino

. New 5-Factor modified frailty index using American college of surgeons NSQIP data. J Am Coll Surg. 2018;226(2):173-181.e8. doi:10.1016/J.JAMCOLLSURG.2017.11.005

Gilbert

Neuburger

Kraindler

, et al. Development and validation of a Hospital Frailty Risk Score focusing on older people in acute care settings using electronic hospital records: An observational study. The Lancet. 2018;391(10132):1775-1782. doi:10.1016/S0140-6736(18)30668-8

Mahanna-Gabrielli

Zhang

Sieber

, et al. Frailty is associated with postoperative delirium but not with postoperative cognitive decline in older noncardiac surgery patients. Anesth Analg. 2020;130(6):1516-1523. doi:10.1213/ANE.0000000000004773

Ethun

Bilen

Jani

Maithel

Ogan

Master

. Frailty and cancer: Implications for oncology surgery, medical oncology, and radiation oncology. CA Cancer J Clin. 2017;67(5):362-377. doi:10.3322/CAAC.21406

Panayi

Orkaby

Sakthivel

, et al. Impact of frailty on outcomes in surgical patients: A systematic review and meta-analysis. Am J Surg. 2019;218(2):393-400. doi:10.1016/J.AMJSURG.2018.11.020

10.

AL-Khamis

Warner

Park

, et al. Modified frailty index predicts early outcomes after colorectal surgery: An ACS-NSQIP study. Colorectal Disease. 2019;21(10):1192-1205. doi:10.1111/codi.14725

11.

Fagard

Leonard

Deschodt

, et al. The impact of frailty on postoperative outcomes in individuals aged 65 and over undergoing elective surgery for colorectal cancer: A systematic review. J Geriatr Oncol. 2016;7(6):479-491. doi:10.1016/J.JGO.2016.06.001

12.

Michaud Maturana

English

Nandakumar

Li Chen

Dvorkin

. The impact of frailty on clinical outcomes in colorectal cancer surgery: A systematic literature review. ANZ J Surg. 2021;91(11):2322-2329. doi:10.1111/ANS.16941

13.

Wheat

Strate

. Trends in hospitalization for diverticulitis and diverticular bleeding in the United States from 2000 to 2010. Clin Gastroenterol Hepatol. 2016;14(1):96. doi:10.1016/J.CGH.2015.03.030

14.

Devon

Vergara-Fernandez

Victor

McLeod

. Colorectal cancer surgery in elderly patients: Presentation, treatment, and outcomes. Dis Colon Rectum. 2009;52(7):1272-1277. doi:10.1007/DCR.0B013E3181A74D2E

15.

Näsvall

Dahlstrand

Löwenmark

Rutegård

Gunnarsson

Strigård

. Quality of life in patients with a permanent stoma after rectal cancer surgery. Qual Life Res. 2017;26(1):55-64. doi:10.1007/S11136-016-1367-6

16.

Wrenn

Cepeda-Benito

Ramos-Valadez

Cataldo

. Patient perceptions and quality of life after colon and rectal surgery: What do patients really want? Dis Colon Rectum. 2018;61(8):971-978. doi:10.1097/DCR.0000000000001078

17.

Kaiser

Israelit

Klaristenfeld

, et al. Morbidity of ostomy takedown. Journal of Gastrointestinal Surgery. 2008;12(3):437-441. doi:10.1007/S11605-007-0457-8/TABLES/2

18.

de Paula

Nemeth

Kiran

Keller

. Predictors of complications from stoma closure in elective colorectal surgery: An assessment from the American College of Surgeons National Surgical Quality Improvement Program (ACSNSQIP). Tech Coloproctol. 2020;24(11):1169-1177. doi:10.1007/s10151-020-02307-5

19.

Shahait

Qadeer

Hasnain

, et al. Hartmann’s reversal outcomes: A VASQIP study. J Gastrointest Surg. 2021;25(2):539-541. doi:10.1007/s11605-020-04807-z

20.

Kang

Yoon

, et al. Analysis of factors affecting reversal of Hartmann’s procedure and post-reversal complications. Sci Rep. 2020;10(1):16820. doi:10.1038/s41598-020-73791-w

21.

Kandagatla

Khawja

Gupta

Rubinfeld

. Readmission after ostomy reversal: Who’s at risk? J Am Coll Surg. 2018;227(4):S114-S115. doi:10.1016/j.jamcollsurg.2018.07.238

22.

Kye

Kim

Cho

. Is it safe the reversal of a diverting stoma during adjuvant chemotherapy in elderly rectal cancer patients? Int J Surg. 2014;12(12):1337-1341. doi:10.1016/j.ijsu.2014.10.018

23.

Bosshardt

. Outcomes of ostomy procedures in patients aged 70 years and older. Arch Surg. 2003;138(10):1077-1082. doi:10.1001/archsurg.138.10.1077

24.

Kooragayala

Lou

Butchy

, et al. Impact of frailty on patient outcomes after Hartmann’s reversal: A NSQIP analysis. Am Surg. 2023. Published online February 14, 2023. doi:10.1177/00031348231156785

25.

Bilimoria

Liu

Paruch

, et al. Development and evaluation of the universal ACS NSQIP surgical risk calculator: A decision aide and informed consent tool for patients and surgeons. J Am Coll Surg. 2013;217(5):833. doi:10.1016/J.JAMCOLLSURG.2013.07.385

26.

Segev

Assaf

Elbaz

, et al. Outcomes of diverting loop ileostomy reversal in the elderly: A case–control study. ANZ J Surg. 2021;91(6):E382-E388. doi:10.1111/ANS.16871

27.

Keller

Khorgami

Swendseid

Khan

Delaney

. Identifying causes for high readmission rates after stoma reversal. Surg Endosc. 2014;28(4):1263-1268. doi:10.1007/S00464-013-3320-X/TABLES/3

28.

Banerjee

Leather

AJM

Rennie

Samano

Gonzalez

Papagrigoriadis

. Feasibility and morbidity of reversal of Hartmann’s. Colorectal Dis. 2005;7(5):454-459. doi:10.1111/J.1463-1318.2005.00862.X

29.

Saur

Davis

Montroni

, et al. The American society of colon and rectal surgeons clinical practice guidelines for the perioperative evaluation and management of frailty among older adults undergoing colorectal surgery. Dis Colon Rectum. 2022;65(4):473-488. doi:10.1097/DCR.0000000000002410

30.

Liu

Kipnis

, et al. Comparison of Electronic Frailty Metrics for Prediction of Adverse Outcomes of Abdominal Surgery. JAMA Surg. Published online March. 2022;16:e220172. doi:10.1001/JAMASURG.2022.0172

Association of Frailty With Post-Operative Outcomes of Older Adults Undergoing Elective Ostomy Reversal

Abstract

Background

Methods

Results

Conclusions

Keywords

Introduction

Methods

Data Source

Patient Selection

Primary Regressor of Interest

Other Predictor Variables

Outcomes

Statistical Analysis

Results

Bivariate Analysis

Univariate Regression Analysis

Multivariate Regression Analysis

Discussion

Footnotes

Declaration of Conflicting Interests

Funding

ORCID iDs

Data Availability Statement

References