SEARCH

SEARCH BY CITATION

Keywords:

  • abdominal surgery;
  • complications;
  • morbidity;
  • mortality

Abstract

  1. Top of page
  2. Abstract
  3. METHOD
  4. VARIABLES
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES

OBJECTIVES: To review the outcomes of patients aged 85 and older after abdominal surgery in terms of mortality, morbidity, and change in residential status and to analyze factors predicting such outcomes.

DESIGN: Retrospective clinical cohort study.

SETTING: A tertiary regional hospital in Victoria, Australia.

PARTICIPANTS: One hundred seventy-nine patients aged 85 and older who had abdominal surgery between 1998 and 2008.

MEASUREMENTS: Mortality, complications (morbidity), and change in residential status.

RESULTS: The patient sample had a mean age of 88.6, a mortality rate of 17.3%, and a morbidity rate of 62.8%. Approximately two-thirds (64%) of all abdominal surgeries were emergency surgeries. Factors predicting mortality included American Society of Anesthesiologists (ASA) score and premorbid residential status. Risk factors predicting severity of complications were ASA score and emergency surgery. Significant factors contributing to change in residential status were ASA score and severity of complications. Age, sex, and number of comorbidities were not significant factors.

CONCLUSION: Patients aged 85 and older experienced mortality rates of 17.3% after abdominal surgery. ASA score and premorbid residential status appear to be more important than age in determining risk for abdominal surgery in older persons.

The developed world is experiencing a demographic transition in which the proportions of people in the oldest age groups are increasing, and the proportions in the youngest age groups are decreasing. In Australia, centenarians are the fastest-growing age segment of the population.1 The Australian Bureau of Statistics projects that the number of people in Australia aged 85 and older will be between 1.6 million and 2.7 million by 2051. It is projected that this age cohort will constitute 6% to 8% of the population at that time, compared with a current percentage of 1.5%. It therefore seems likely that Australia will experience a similar increase in the incidence of geriatric surgery over the next 40 to 50 years. A retrospective study examining the outcomes and risk factors of patients aged 85 and older after abdominal surgery was conducted. Factors that contributed to death, complications, and change in residential status were determined.

METHOD

  1. Top of page
  2. Abstract
  3. METHOD
  4. VARIABLES
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES

The study examined 179 patients aged 85 and older who had abdominal surgery at Geelong Hospital (a tertiary hospital) between 1998 and 2008. Data on patients aged 90 and older (n=62) were consecutively collected over 10 years, and data on patients aged 85 to 90 (n=117) were consecutively collected over 5 years (2003–2008). Data for nonagenarians were collected over a longer period of time to obtain a more-even spread of ages in the sample. Patients who had undergone urological, gynecological, vascular, or endoscopic procedures were excluded. Data were collected using an electronic hospital database to include age, abdominal surgery, and year of surgery. All of the other data were then collated from patient's existing medical records.

VARIABLES

  1. Top of page
  2. Abstract
  3. METHOD
  4. VARIABLES
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES

Four variables used in the analysis require discussion. In 1963, the American Society of Anesthesiologists (ASA) adopted a five-category physical status classification system for assessing a patient before surgery.2

  • I. A normal healthy patient.
  • II. A patient with mild systemic disease.
  • III. A patient with severe systemic disease.
  • IV. A patient with severe systemic disease that is a constant threat to life.
  • V. A moribund patient who is not expected to survive without the operation.

In this study, the anesthetist assigned ASA scores at the time of surgery, and the nursing staff recorded them in an electronic format in the medical database. Data collected in such format were then manually crosschecked with the patient record to ensure validity and accurate documentation.

Second, a postoperative complication is defined as any deviation from the normal expected postoperative course. The grading system used in this study is a modification of classifications proposed previously.3,4 This four-grade system is standard in Australia and is recommended by The Royal Australasian College of Surgeons.5 The grade is based on the therapy used to treat the complications, and it thus prevents a down rating of negative outcomes especially in a retrospective analysis such as this one. In addition, it was useful to have a numerical value for the regression analysis performed below. The primary author (NM), who was not involved in any of the observed surgeries, graded the complications, which were based on the medical records. To reduce bias, the person grading the complication was unaware of the patient's age.

Third, change in residential status (CiRS) was measured by recording progressive deterioration in living status: home to rehabilitation, home to aged care, aged care to nursing home, and home to nursing home. Aged care was assumed to be a low to medium level of care, as opposed to a nursing home, which is always a high level of care.

Fourth, comorbidities are defined as the presence of one or more diseases in addition to the primary disease or disorder. Examples include ischemic heart disease, current or ex-smoker, alcohol abuse, high blood pressure, and diabetes mellitus. Patients' comorbidities were converted to discrete numerical data, regardless of type or severity. Each disease was classified as one entity, as recommended previously.6 This method will account only for overall burden of disease and fails to include prognostic differences of each disease. Other studies7,8 have analyzed comorbidities in the same way, although Charlson et al.9 gave each comorbidity a weighting according to its severity. A weighted score would be more difficult to apply because, in many hospitals, details of patient's comorbidities are inadequately documented.

Statistical Analysis

To examine the factors that explain postoperative outcomes, three statistical models were estimated. First, a multivariate probit model was used to model the probability of death as a function of age, ASA score, premorbid function, comorbidities, variables for admission type (emergency=1, elective=0), sex (female=1, male=0), and operator type (consultant=1, registrar=0). The postoperative complication severity and postoperative CiRS variables were modelled using ordered probit models. Both models used the same predictors as the probability of death model except for the CiRS regression, which also included complication severity. (Complication severity was not included in the probability-of-death model, because it perfectly predicted the death outcome, thereby violating the regression assumptions.) A full model was initially estimated that included all of the predictor variables. The model reported only included the significant variables.

RESULTS

  1. Top of page
  2. Abstract
  3. METHOD
  4. VARIABLES
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES

Patient Demographics

The study population comprised 107 women and 72 men with a mean age of 88.6. ASA III was the most common score (Table 1). The majority of patients (149; 83.2%) were admitted from home, 20 (11.2%) from aged care, and 10 (5.6%) from a nursing home. The average number of comorbidities was 2.5, with the most common comorbidity being cardiovascular disease.

Table 1. Patient Demographics
CharacteristicValue
  • *

    The percentages for comorbidities add to more than 100% because one patient could have more than one condition.

Sex, n (%)
 Male72 (40.2)
 Female107 (59.8)
Age, average (range)88.6 (85–97)
American Society of Anesthesiologist score, n (%)
 I3 (1.7)
 II33 (18.4)
 III83 (46.4)
 IV56 (31.3)
 V4 (2.2)
Premorbid residential status, n (%)
 Home149 (83.2)
 Aged care20 (11.2)
 Nursing home10 (5.6)
Number of comorbidities, average2.5
Comorbidities, n (%)*
 Cardiac disease70 (39.1)
 Hypertension66 (36.9)
 Asthma or chronic obstructive pulmonary disease25 (14.0)
 Dementia13 (7.3)
 Atrial fibrillation34 (19.0)
 Gastroesophageal reflux29 (16.2)
 Diabetes mellitus20 (11.2)
 Ex-smoker or smoker23 (12.8)
Number of days in the hospital, average13.7

Operation Demographics

The majority of abdominal surgeries performed were emergencies (64%). Hernia repairs and colorectal procedures were the majority of elective surgeries (Table 2). Consultant surgeons performed 70% of abdominal operations. Colorectal operations were one of the most common abdominal surgeries performed (31%), a likely testament to the increasing risk of colorectal cancer in this age group. The results showed that 33% (10/30) of laparoscopic cholecystectomies were converted to open. The majority of those converted to open were emergency cases (70%).

Table 2. Surgical Variables
Variablen (%)
Operation status
 Emergency115 (64.2)
 Elective64 (35.8)
Primary operator seniority
 Consultant126 (70.4)
 Registrar53 (29.6)
Operation type
 Laparoscopic cholecystectomy20 (11.2)
 Laparoscopic cholecystectomy converted to open10 (5.6)
 Open cholecystectomy3 (1.7)
 Laparotomy or division of adhesion with or without small bowel resection45 (25.1)
 Colorectal56 (31.3)
 Appendicectomy4 (2.2)
 Hernia repair24 (13.4)
 Oversewing of perforated ulcer5 (2.8)
 Rectopexy2 (1.1)
 Other10 (5.6)

Patient Outcomes

The inpatient postoperative mortality rate was 17.3%, with the majority of deaths in emergency cases (Table 3). The postoperative morbidity or complication rate was 62.8% (excluding patients who died), with 59.1% experiencing minor complications only (Grade I). The most common complications were acute pulmonary edema and rapid atrial fibrillation. After abdominal surgery, 64.2% of patients had no change in their residential status. Overall, 28.4% required rehabilitation before being discharged home or to aged care. Only five patients (3%) who were admitted from home were discharged to a nursing home postoperatively. The average inpatient stay was 13.7 days.

Table 3. Outcomes
Characteristicsn (%)
  • *

    The percentages do not add to 62.8% because patients with complications may have had more than one.

Deaths
 Postoperative (of 179 cases)31 (17.3)
 Emergency (of 115 cases)26 (22.6)
 Elective (of 64 cases)5 (7.8)
Complications
 Yes93 (62.8)
 No55 (37.2)
Complication severity
 Grade 0: no complications55 (37.2)
 Grade 1: condition that does not delay discharge and has little effect on patient's well-being (e.g., urinary tract infection)55 (37.2)
 Grade 2: condition that prolongs stay and causes significant morbidity (e.g., pneumonia, ileus)25 (16.9)
 Grade 3: condition that require a radiological or surgical intervention to be managed (e.g., anastomotic leak)13 (8.8)
Complication*
 Wound infection7 (3.9)
 Pulmonary edema18 (10.1)
 Chest infection10 (5.6)
 Myocardial infarction6 (3.4)
 Atrial fibrillation18 (10.1)
 Confusion10 (5.6)
 Transient ischemic attack3 (1.7)
 Urinary retention11 (6.1)
 Urinary tract infection7 (3.9)
 Surgical complication requiring intervention13 (7.3)
 Other14 (7.8)
Residential status
 Change53 (35.8)
 No change95 (64.2)
Severity of change in residential status
 Home to rehabilitation42 (28.4)
 Home to aged care3 (2.0)
 Aged care to nursing home3 (2.0)
 Home to nursing home5 (3.4)

Predicting Outcomes

The results illustrate that ASA score (odds ratio (OR)=2.06, 95% confidence interval (CI)=1.44–2.95, P<.001) and premorbid residential status (OR=1.50, 95% CI=1.02–2.21, P=.04) are statistically significant predictors of mortality. The worse the premorbid function (0=home, 1=aged care, 2=nursing home) and the higher the ASA score, the greater the probability of postoperative mortality. The results for the second regression indicate that the higher the ASA score (OR=1.65, 95% CI=1.25–2.17, P<.001) and emergency status (OR=1.55, 95% CI=1.01–2.40, P=.047) result in a higher likelihood of greater postoperative complication severity. Finally, the factors most predictive of CiRS were ASA score (OR=1.60, 95% CI=1.14–2.24, P=.007), emergency status (OR=1.58, 95% CI=0.94–2.66, P=.08), and complication severity (OR=1.30, 95% CI=1.03–1.65, P=.03).

Age, sex, number of comorbidites, and operator (consultant v. registrar) were not found to be statistically significant risk factors for mortality, postoperative complication severity, or change in residential status.

DISCUSSION

  1. Top of page
  2. Abstract
  3. METHOD
  4. VARIABLES
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES

Mortality associated with abdominal surgery in this series of patients was 17.3%. A literature review of mortality rates in older patients identified rates varying from as low as 8%10 to as high as 32%.11 Although the mortality rate in the current study is comparable with that reported in literature, it is relatively high. In most studies, emergency surgery was associated with higher mortality rates,10,12,13 although this was not demonstrated in this study. A previous study similarly showed no statistically significant difference between emergency and elective surgery with respect to mortality, despite a higher rate of complications after emergency surgery.14 Older age had no correlation with any outcome, including mortality or morbidity. This is consistent with findings in other studies.11,15

The risk factors most predictive of mortality in the current study population were found to be ASA score and premorbid residential status. Premorbid status is likely to be a surrogate marker of frailty of the patient and thus a useful marker for assessing risk. A study of 116 nonagenarians similarly demonstrated that the postoperative survival rate of patients admitted from a nursing home (median years of survival 0.7) was significantly worse than that of patients admitted from a family residence (median years of survival 1.59; P=.04).16 The current study examined only in-hospital, same-admission mortality. ASA score has invariably and robustly been shown to be a reliable predictor of postoperative mortality in other studies in older populations,15,17–20 although ASA score can be biased, because it is open to subjectivity and manipulation.21 Anesthetists are expected to classify patients with reference to their physical status alone and not be influenced by knowledge of proposed surgery.2

Higher rates of conversions to open cholecystectomies were noted in the current study. A previous study showed a higher rate of conversion to laparotomy in an older population (≥65) than in a younger age group.22 Such a high rate of conversion to open in the elderly patients studied in the current study supports the argument made by the previous study that this is likely to be due to greater acuity and chronic right upper quadrant inflammation in this population. In the previous study, acute and chronic cholecystitis were the operative diagnoses in a greater proportion of the elderly group, as was the incidence of prior upper abdominal surgery.

The complication rate in the current study was 62.8%, with most common complications being pulmonary edema and rapid atrial fibrillation. A morbidity rate of up to 57% has been reported in the literature for this group of patients.13,14 Factors that predicted the severity of complications were ASA score and emergency surgery. A previous study used a similar grading system for complication and showed that ASA score was a significant predictive factor for the severity of postoperative complications.19 Other studies have also shown higher morbidity rates associated with emergency surgery.13,14 Such an outcome may be because, in an emergency setting, preparation of the patient can be compromised because of time limitations and most patients are unlikely to be in optimal medical status before surgery. Although the analysis in the current study failed to show emergency surgery as a risk factor for mortality, a clear significant relationship was seen when predicting complication severity.

The most common comorbidity was cardiovascular disease; in one study, this was one of the most common coexisting diseases,14 and in another, cardiogenic shock was a leading cause of mortality.20 The number of comorbidities were not found to be a significant predictor in our study, although another study showed that two or more comorbidities was a significant predictor of mortality and morbidity in elective colorectal surgery.8

The majority of patients in the current study had no change in residential status (64.2%). Of those who experienced a change, 79.2% were discharged to rehabilitation with a possibility of requiring further placement if rehabilitation was unsuccessful. The detail of definite residential status after rehabilitation was not available for analysis. A minority required discharge directly to a nursing home. A study of a large cohort of patients (N=3,736) observing the effect of age on outcomes after pancreatic surgery showed that older patients were less likely to be discharged home and more likely to require care at an inpatient nursing or acute care facility at the time of discharge.23 It showed that 93.8% of patients aged 60 to 69 were discharged home with or without home help, compared with 61.8% of those aged 80 and older. The factors most predictive of a change in the current study were ASA score and severity of complications. Emergency surgery was only significant at a 10% level of significance.

ASA score is undoubtedly the most important risk factor for all three outcomes analyzed. The premorbid functioning of patients would reflect their frailty and thus the extent of their health status. It is therefore not surprising that this was also a significant variable predicting mortality. The insignificance of age is an important finding of this study. Within the age range of 85 to 97, the actual age of the patient did not have a statistically significant effect on outcomes from abdominal surgery.

Other insignificant factors included sex and seniority of operator. A previous study similarly showed little variation in mortality or morbidity according to seniority of the surgeon,17 although the results may be biased because consultant surgeons may have operated on more-complex cases. The literature, in general, suggests that there is no difference in patient outcome regarding the seniority of the surgeon when a supervised trainee or a surgeon performs the surgery.24,25 In Geelong Hospital, it would be standard practice for all trainees to be directly supervised for major surgeries.

This study was limited by being a retrospective medical record review. Furthermore, the number of elderly patients admitted during the same time period with a surgical problem that were treated conservatively is unknown. Thus the sample population studied may have overrepresented the “fitter” group of elderly presenting with a surgical problem and thus underestimated mortality and morbidity rates. It was also not possible to verify from medical records the final definitive discharge destination of patients who were admitted to a rehabilitation center.

CONCLUSION

  1. Top of page
  2. Abstract
  3. METHOD
  4. VARIABLES
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES

This study provides a comprehensive analysis of pertinent outcomes after abdominal surgery in a sample of elderly patients. The factors predicting patient outcomes were also analyzed, thus aiding in the decision-making process for patients and surgeons. This study found that patients aged 85 and older had mortality rates of 17.3%. It also showed that mortality, complication severity, and the possibility of a change in residential status postoperatively could be predicted according to patient ASA score. Premorbid functional status is also an important variable predicting mortality. Although it can be concluded that the results are not enough to discourage operating and that age is not a contraindication to surgery, it is not surprising that mortality and morbidity rates are higher in elderly patients. Thus the decision to operate on elderly people should include informed consent and be based on the predicted benefit from surgery, with due consideration given to the patient's preoperative risk factors.

ACKNOWLEDGMENTS

  1. Top of page
  2. Abstract
  3. METHOD
  4. VARIABLES
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES

Conflict of Interest: The editor in chief has reviewed the conflict of interest checklist provided by the authors and has determined that the authors have no financial or any other kind of personal conflicts with this paper.

Author Contributions: Naseem Mirbagheri: data collection, interpretation of data, and preparation of manuscript. Jonathan Dark: data collection, interpretation of data, statistical analysis, and preparation of manuscript. David Watters: study design, interpretation of data, and review of manuscript.

Sponsor's Role: None.

REFERENCES

  1. Top of page
  2. Abstract
  3. METHOD
  4. VARIABLES
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES