0
We're unable to sign you in at this time. Please try again in a few minutes.
Retry
We were able to sign you in, but your subscription(s) could not be found. Please try again in a few minutes.
Retry
There may be a problem with your account. Please contact the AMA Service Center to resolve this issue.
Contact the AMA Service Center:
Telephone: 1 (800) 262-2350 or 1 (312) 670-7827  *   Email: subscriptions@jamanetwork.com
Error Message ......
Original Article |

Adverse Outcomes of Geriatric Patients Undergoing Abdominal Surgery Who Are at High Risk for Delirium FREE

Sabha Ganai, MD; K. Francis Lee, MD, MPH; Andrea Merrill, BS; Maggie H. Lee, MS; Sandra Bellantonio, MD; Maura Brennan, MD; Peter Lindenauer, MD, MS
[+] Author Affiliations

Author Affiliations: Departments of Surgery (Drs Ganai and Lee and Ms Lee) and Medicine (Ms Merrill and Drs Bellantonio, Brennan, and Lindenauer), Baystate Medical Center, Tufts University School of Medicine, Springfield, Massachusetts.


Arch Surg. 2007;142(11):1072-1078. doi:10.1001/archsurg.142.11.1072.
Text Size: A A A
Published online

Hypothesis  Among geriatric patients undergoing abdominal surgery who are at high risk for in-hospital delirium, clinical factors associated with delirium correlate with adverse outcomes.

Design  Retrospective case series study.

Setting  University-affiliated referral hospital.

Patients  Among 228 consecutive patients 70 years or older who underwent major abdominal surgery from September 1, 2002, through December 31, 2003, 89 patients with risk factors for delirium were included in the study.

Main Outcome Measures  Preoperative, intraoperative, and postoperative clinical factors known to affect the incidence of in-hospital delirium were tested for correlation with adverse outcomes. Incidence of delirium, mortality, and prolonged length of stay (LOS) of 14 days or longer were evaluated as adverse outcomes.

Results  Postoperative delirium occurred in 60%, death in 20%, and prolonged LOS in 32% of patients. Multivariate analysis identified independent predictors of adverse outcomes. Poor preoperative functional and nutritional status correlated with postoperative delirium and mortality. Inadequate postoperative glycemic control also correlated with mortality. Complications in 2 or more organ systems and postoperative hypoalbuminemia (albumin level <3.0 mg/dL[<.003 g/dL; to convert to grams per liter, multiply by 10]) correlated with prolonged LOS. Suboptimal care was identified in the following clinical areas: use of precipitative medications, prolonged bedrest, uncontrolled pain, hypoxia, and glycemic control.

Conclusions  In a subset of geriatric patients undergoing abdominal surgery who are at high risk for in-hospital delirium, adverse outcomes correlated only with key clinical variables, such as hyperglycemia and poor nutritional and functional states. A high incidence of suboptimal care was observed in several clinical areas, suggesting opportunities for intervention.

Figures in this Article

Delirium is present in 22% to 31% of hospitalized elderly patients1,2 and in 37% of postoperative geriatric patients, in some cases as many as 74%.3 Among patients undergoing cardiac, general, orthopedic, and vascular procedures, postoperative delirium has been found to correlate with a variety of preoperative, intraoperative, and postoperative clinical variables.47 The patients in these studies were not stratified according to their risks for delirium, and the findings apply to patient populations at low risk and high risk for delirium. Evidence has suggested that strategies for reducing postoperative delirium may be different for patients at various risk levels.8

To our knowledge, no studies have evaluated geriatric patients undergoing major abdominal surgery who are identified at admission as high risk for postoperative delirium. With 60% of all general surgical procedures being performed on patients older than 65 years,9 prevention and treatment of delirium among patients who are at high risk of delirium are critical to affect both cost and outcome.

Inouye et al10 have created a validated prediction model for delirium in hospitalized, elderly patients based on 4 risk factors at admission: (1) severe illness, (2) visual impairment, (3) cognitive impairment, and (4) serum urea nitrogen to creatinine (SUN:Cr) ratio of 18 or higher. The model revealed an 83% incidence of delirium among patients with 3 or 4 risk factors compared with 9% among patients with no risk factors. A significant difference was demonstrated in mortality and nursing home placement outcomes based on risk factors at admission.

Using the model of Inouye and colleagues, we identified a group of geriatric patients undergoing abdominal surgery who were determined at admission to be at high risk for postoperative delirium. We reviewed their clinical characteristics and outcomes to study the hypothesis that adverse outcomes would be frequent among these high-risk cohorts and would correlate with key clinical factors previously known to be associated with the development of delirium.

STUDY INCLUSION CRITERIA

The study population consisted of patients who were admitted to a 650-bed teaching hospital from September 1, 2002, through December 31, 2003. A retrospective review of hospital medical records was conducted for all patients 70 years or older who underwent abdominal surgery. Inclusion criteria were applied, identifying patients with 3 or 4 of the following risk factors at admission for in-hospital delirium previously validated by Inouye et al10: (1) severe illness, (2) visual impairment, (3) cognitive impairment, and (4) dehydration as determined by a SUN:Cr ratio of 18 or higher. In our study, we modified the original validation model by requiring severe illness as an inclusion criterion, defined as having diseases that needed major abdominal surgical procedures such as open abdominal explorations for bowel perforation, obstruction, bleeding, ischemia, infection, and cancer. We excluded less serious diseases that required procedures such as uncomplicated hernia repair or cholecystectomy. Visual impairment was identified in the medical records by a nurse's documentation of the patient's use of corrective lenses, presence of cataracts, diabetic retinopathy, or macular degeneration. Cognitive impairment was assessed by documentation of a prior history of dementia, Alzheimer disease, or significant recent memory loss.

PATIENT CHARACTERISTICS

Clinical data were abstracted from hospital medical records of patients who met the inclusion criteria. Baseline patient characteristics were recorded, including age, sex, language, presence of advance directive, and family participation in care. Admission source and elective or emergency nature of surgery were noted. For each patient, the medical record was reviewed by at least 2 separate investigators (S.G., K.F.L., and A.M.) for concurrence to minimize investigator bias.

CLINICAL FACTORS ASSOCIATED WITH IN-HOSPITAL DELIRIUM

On the basis of previous reports,4,1114 the following preoperative, intraoperative, and postoperative clinical variables were considered as factors associated with in-hospital delirium. They were tested for possible correlation with adverse outcomes.

Preoperative factors included poor nutritional status, poor functional status, and a history of diabetes mellitus. Poor nutritional status was defined as a serum albumin level less than 3.0 mg/dL (<.003 g/dL; to convert to grams per liter, multiply by 10), precipitous weight loss, or physical description of malnourishment and emaciation. Poor preoperative functional status was determined by documentation of a prior inability to independently care for self, with 1 or more deficits in various activities of daily living (bathing, dressing, toileting, transfer, continence, or feeding).15 History of diabetes was confirmed by documentation of a prior oral hypoglycemic agent or insulin use. History of alcohol intake was also noted.

Intraoperative factors were assessed by review of anesthesia and operative notes for the occurrence of blood transfusion during the procedure, excessive blood loss, excessive fluid administration, and intraoperative hypotension. Excessive blood loss was defined as intraoperative blood loss greater than 400 mL. Excessive fluid administration was defined as intravenous crystalloid and colloid intake greater than 5 L. Intraoperative hypotension was defined as any documented mean arterial pressure less than or equal to 60 mm Hg or the prolonged use of vasoactive substances (eg, norepinephrine and phenylephrine).

Postoperative factors were assessed for presence in the first 72 hours after surgery. They included documentation of physical restraint, bladder catheterization, administration of more than 4 new medications, the use of precipitative medications, prolonged bed rest, poor fluid status (SUN:Cr ratio ≥18 or hourly urine output <0.5 mL/kg), poor glycemic control (serum glucose level >150 mg/dL; to convert to millimoles per liter, multiply by 0.0555), hypoxia (arterial oxygen saturation <95%), anemia (hematocrit <30%), uncontrolled pain, metabolic derangement, and hypoalbuminemia (albumin level <3.0 mg/dL). Potentially precipitative medications included meperidine, hydroxyzine, benzodiazepines, diphenhydramine, tricyclic antidepressants, muscle relaxants, and barbiturates.11,14 Metabolic derangement was defined as having one of the following serum abnormalities: pH less than 7.34 or greater than 7.55, serum sodium level greater than 150 or less than 130 mEq/L (to convert to millimoles per liter, multiply by 1.0), or serum potassium level greater than 5.0 or less than 3.0 mEq/L (to convert to millimoles per liter, multiply by 1.0). Uncontrolled pain was defined as a pain score greater than 5 of 10 as documented by nursing staff along with assessment of vital signs.

In-hospital morbidity was assessed by reviewing complications that occurred throughout the hospital course. Complications were categorized by the associated organ system as follows: neurologic (stroke or seizures), cardiac (myocardial infarction, acute heart failure, or arrhythmias), respiratory (acute respiratory distress syndrome, pneumonia, or pleural effusion), gastrointestinal or hepatobiliary (hepatic insufficiency, jaundice, or prolonged ileus), renal (acute renal failure or electrolyte disturbances), infectious (bacterial sepsis, fungemia, or urinary tract infection), vascular (arterial or venous embolism), and wound (abdominal wall dehiscence or wound infection). The total number of organ system complications and morbidity rates were tabulated per patient.

SUBOPTIMAL CARE IN CLINICAL AREAS ASSOCIATED WITH DELIRIUM

Suboptimal care was defined as unexplained failure to recognize and/or treat a clinical issue. Continued use of physical restraint was considered suboptimal care if there was no clinical need for restraint for the protection of catheters, tubes (eg, endotracheal), or monitoring devices. Prolonged use of a bladder catheter was considered suboptimal care if a catheter was present despite adequate urine output, discontinuation of use of epidural anesthesia, or full ambulatory status. Prolonged bed rest was considered inappropriate in extubated patients if unexplained by a physician or a physical therapist.

If more than 4 new medications were administered, reasons for their use were evaluated. Appropriate uses for prophylactic antibiotics, diuretics, analgesics, or sedatives were excluded. Suboptimal care was noted when the medication list was not adjusted as clinical needs dissipated. When precipitative medications were given for analgesia or sedation, suboptimal care was noted if use of the drug could have been avoided by an alternate medication not associated with delirium. Pain was considered suboptimally managed if a pain score of greater than 5 was recorded during 8 hours without adequate analgesia to decrease the pain score below 5 of 10. Suboptimal care was also considered to be present if the pain score was not recorded at all by the physicians or nurses.

Malnutrition was considered suboptimally managed if documentation of nutritional status was absent, institution of nutritional support (parenteral or enteral) based on physical examination was delayed, or a nutrition consultation was lacking. Poor postoperative fluid status was considered suboptimally managed if there was no documentation to suggest aggressive fluid resuscitation in the setting of clinically evident dehydration, such as an hourly urine output of less than 0.5 mL/kg, in the absence of heart failure. Metabolic derangements were considered suboptimally managed if appropriate electrolyte repletion or fluid intervention was not undertaken in response to abnormal laboratory test results. Glycemic control was considered suboptimally managed if serum glucose levels remained greater than 150 mg/dL and no measures were taken to administer insulin. Hypoxic episodes were considered suboptimally managed if supplemental oxygen was not administered in response to recorded arterial oxygen saturation of less than 95%. Hematocrit less than 30% was considered suboptimally managed if there was no documentation that reflected recognition of anemia.

OUTCOME END POINTS

Adverse outcome end points included occurrence of delirium, in-hospital mortality, and prolonged hospital length of stay (LOS). Presence of delirium was determined by the physician's or nurse's notation of symptoms consistent with the definition described in the Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition, Text Revision) (DSM-IV-TR).16 The symptoms included new onset of patient disorientation, diminished attention to external stimuli, disordered communication, confusion, agitation, or other indications of severe fluctuations in mental status. Mortality was defined as death before discharge from the hospital. Prolonged LOS was defined as 14 or more days of hospitalization.

STATISTICAL ANALYSIS AND STUDY DESIGN

The study methods received institutional review board approval. Statistical analysis was conducted with the Epi Info 3.2.2 software package (Centers for Disease Control and Prevention, Atlanta, Georgia). Statistical significance was determined with an α of .05 using 2-tailed P values. Differences in proportions were evaluated with the Mantel-Haenszel χ2 test. The Mann-Whitney rank sum test was used to study differences in means. Means and relative risks (RRs) are indicated with 95% confidence intervals (CIs) in parentheses.

After performance of univariate analysis, clinical factors that achieved statistical significance were considered for inclusion in a stepwise unconditional multiple logistic regression analysis model. The Newton-Rhapson method17 was used to determine maximum likelihood estimates of the regression coefficient (β), the standard error of coefficient (SE), and the significance level (P) for each factor. The normal statistic (z) for each factor is the ratio between the regression coefficient and the SE and indicates the relative importance of that factor in determining outcome. The adjusted odds ratio for the variable represents the outcome for the patients with the factor present compared with the factor absent after adjusting for other factors. Clinical factors were entered into logistic regression models, with variables with a minimum normal statistic removed sequentially until statistically significant independent variables of association remained.

A total of 228 patients 70 years or older who underwent major abdominal surgery during the study period were evaluated for inclusion in the study. Of the 228 patients, 89 (39%) had 3 or more risk factors to meet the admission inclusion criteria for being at high risk for in-hospital delirium (mean age, 79 years; 95% CI, 77-80 years). Table 1 lists the clinical profiles of the patients. The Figure displays patients' admission sources and discharge destinations. Patients who developed postoperative delirium were less likely to be discharged to home than those who did not (31% vs 72%; P <.001). The indications for major abdominal surgical procedures are listed in Table 2. The mean hospital LOS was 12 days (95% CI, 10-13 days). Thirty-six of 89 patients (40%) were admitted to the intensive care unit (ICU) postoperatively, with a mean ICU stay of 6 days (95% CI, 4-8 days). No elective cases were associated with an ICU admission.

Table Graphic Jump LocationTable 1. Baseline Characteristics of 89 Patients at High Risk for Delirium
Place holder to copy figure label and caption
Figure.

Admission source of all 89 geriatric patients undergoing abdominal surgery and discharge destination of the 71 survivors.

Graphic Jump Location
Table Graphic Jump LocationTable 2. Indications for Abdominal Surgery for 89 Patients at High Risk for Delirium

The overall incidence of delirium, mortality, and prolonged LOS was 60%, 20%, and 32%, respectively. Patients who required ICU admission had 50% mortality, of which all cases were associated with emergency operations. Table 3 indicates postoperative morbidity, involving a mean of 1.6 (95% CI, 1.2-1.9) organ systems per patient overall. Patients with delirium had mean of 1.9 organ systems with morbidity (95% CI, 1.5-2.4) compared with 1.0 (95% CI, 0.7-1.4) for those without delirium (P =.01).

Table Graphic Jump LocationTable 3. Morbidity by Organ System Among 89 Patients at High Risk for Delirium

No statistically significant difference was seen in age among patients who survived vs died (P = .12). A small but statistically significant difference in age was seen between the group that developed delirium (81 years; 95% CI, 79-82 years) vs those who did not (77 years; 95% CI, 75-79 years; P = .005). A LOS of 14 days or longer was seen in 28 patients (32%), of whom 11 (38%) had a delay in discharge after full restoration of oral dietary intake and bowel function. Primary reasons for discharge delay included management of complications in 9 (80%) and anticipation of institutional placement in 2 (20%). Patients with delirium were more likely to have an LOS of 14 days or longer than those without delirium (RR, 2.1; 95% CI, 1.02-4.5; P = .03). The presence of delirium correlated with increased risk of death (RR, 1.5; 95% CI, 1.3-1.8; P <.001). In addition, among patients who survived their hospitalization, patients who developed postoperative delirium were less likely to be discharged to home (RR, 0.4; 95% CI, 0.3-0.7; P < .001).

Table 4 summarizes various clinical factors previously described in the literature to be associated with delirium. Suboptimal care was noted in most patients affected by precipitative medications, prolonged bed rest, poor glycemic control, and hypoxia. Uncontrolled pain and anemia were also inadequately addressed in nearly half of the patients affected. Of the 89 patients studied, only 40 (45%) had albumin levels recorded within the first 3 postoperative days, with a mean albumin level of 2.5 mg/dL (95% CI, 2.3-2.7 mg/dL) (.0025 g/dL [95% CI, .0023-.0027 g/dL]. In contrast, suboptimal care was relatively infrequent in the areas of postoperative fluid status and metabolic derangements.

Table Graphic Jump LocationTable 4. Postoperative Clinical Factors and Incidence of Suboptimal Care Among 89 Patients at High Risk for Delirium

Table 5 gives the results of univariate analysis of clinical factors presumably associated with adverse outcomes. Table 6 gives the results of multiple logistic regression for variables associated with adverse outcomes. Preoperative nutritional and functional status were independent predictors of delirium. No intraoperative or postoperative clinical factors were associated with delirium after multivariate analysis. Poor glycemic control, preoperative nutritional status, and preoperative functional status were independent predictors of death. Hypoalbuminemia (albumin level <3.0 mg/dL [<.003 g/dL]) and morbidity (>2 organ systems involved) were independent predictors of prolonged LOS (≥14 days).

Table Graphic Jump LocationTable 5. Univariate Analyses of Association of Clinical Factors With Adverse Outcomes
Table Graphic Jump LocationTable 6. Adjusted ORs for Independent Predictors of Adverse Outcomes Using Logistic Regression

In 1341 patients undergoing elective noncardiac procedures, Marcantanio et al4 reported that low-risk patients had a less than 1% incidence of postoperative delirium, whereas high-risk patients had an incidence of 45%. On the basis of a prediction model, postoperative delirium was predicted in high-risk patients with an accuracy of 70%.7 In addition to polypharmacy and dose-dependent correlation with age, various perioperative factors have been implicated in the development of delirium,3,4,7,1114 including ineffective pain control and sleep cycle disturbances.18,19

The logical conclusion of these findings is that postoperative delirium depends on several preoperative, intraoperative, and postoperative clinical factors, with the potential for reduction in incidence by optimizing care by focusing on the subgroup of high-risk patients. Because postoperative delirium is believed to be linked to adverse outcomes, optimal management of the associated clinical factors may lead to improved clinical outcomes, such as mortality and LOS.

The present study was conducted as a retrospective exploration to test this rationale, specifically for geriatric patients undergoing abdominal surgery who are identifiable at admission as being at high risk for postoperative delirium. To our knowledge, this particular patient subset has not been previously studied in the literature. Certainly, it would have been ideal to conduct a prospective randomized trial to obtain a higher level of evidence. However, because of resource constraints, a carefully designed retrospective analysis was performed as a pilot exploration to answer, with some caution, the foregoing question.

Our patient population yielded a 60% incidence of postoperative delirium, which is higher than the 45% predicted for the high-risk patients undergoing noncardiac surgery in the study by Marcantonio et al.4 Our definition of high risk was borrowed from the study by Inouye et al10 on a sample of nonsurgical patients. Their criterion, severe illness, was construed as equivalent to and therefore replaceable in our study by the stress of major abdominal surgery. Our 60% delirium rate was between the 45% reported by Marcantonio etal and the predicted incidence of 83% for the high-risk nonsurgical patients reported by Inouye et al.10

The postoperative mortality rates in our study were consistent with the findings of Hamel et al20 from the Veterans Administration National Surgical Quality Improvement Project (VA-NSQIP). In the VA-NSQIP database of elderly patients, patients with more than 48 hours of ventilatory assistance, acute renal failure, and systemic sepsis had mortality rates of 38.5%, 52.1%, and 45.2%, respectively, which is similar to the 50% mortality rate among our patients who required postoperative ICU care. Our mortality rates were also consistent with a recent review21 of the literature on the outcomes of critically ill, geriatric postoperative patients.

In univariate analysis, our patients undergoing abdominal surgery who were at high risk for in-hospital delirium showed clinical correlates of postoperative delirium similar to those of the patients undergoing mixed surgical and vascular procedures.4 However, in multivariate analysis using stepwise logistic regression, only the preoperative clinical factors (poor functional and nutritional status) remained as independent clinical correlates.

In analysis of mortality, postoperative clinical variables such as inadequate fluid hydration and hypoalbuminemia — both significant associated factors in univariate analysis — dropped out after logistic regression. Once again, preoperative functional and nutritional status remained as independent clinical correlates of mortality, along with a sole postoperative clinical variable, namely, inadequate glycemic control. These findings conflict with some of the previous reports on postoperative delirium4,7,11,12,18,22 in which additional intraoperative and postoperative clinical factors were reported to significantly influence the occurrence of postoperative delirium.

One explanation may be the difference in the stratification of risks for postoperative delirium. In previous reports4,1114 based on unstratified subsets of patients with a wide range of risks for postoperative delirium, intraoperative and postoperative clinical factors may exhibit heterogeneity with differences that affected the clinical outcome. In our study, all or nearly all of the patients in our subset had severe illness that required major abdominal surgery (100%), emergency surgery (81%), and acute dehydration leading to azotemia (96%). Delirium notwithstanding, these clinical characteristics describe a narrowly defined, acutely ill patient population, portending a generally adverse postoperative course in which many clinical factors may not have made any difference in the outcome.

The only independent correlates of adverse outcomes during intraoperative and postoperative periods were inadequate glycemic control for mortality and postoperative hypoalbuminemia and morbidity (involving >2 organ systems) for prolonged LOS. These clinical factors have been described in the literature more in association with critical care and surgical outcomes than in relation to postoperative delirium. Van den Berghe et al23 have reported level I evidence that intensive insulin therapy for hyperglycemia in critically ill patients reduces mortality and morbidity. Khuri et al24 have also reported level I evidence from the VA-NSQIP database that hypoalbuminemia is first among clinical variables that correlate with adverse postsurgical outcomes.

The findings in the present study cast doubt on the role of conventional interventions (eg, multicomponent management of putative risk factors25,26) designed to prevent postoperative delirium, particularly among patients undergoing abdominal surgery who are at high risk for in-hospital delirium. This suggestion is also consistent with the data reported in the literature. A systematic review26 of 8 studies that involved patients undergoing surgery concluded that interventions to prevent delirium were only modestly effective. Among patients with chronic cognitive impairment, a recent review27 in the Cochrane database determined that evidence was inadequate to warrant a clinical guideline for a multidisciplinary team intervention. Even the clinical investigators who have pioneered the multicomponent intervention to reduce in-hospital delirium pointed out that it was not found to be cost-effective in patients at high risk for delirium.8 The intervention was found to be cost-effective only in patients with an intermediate risk for delirium.

The most effective way to influence adverse outcomes may be to focus, if possible in elective cases, on preoperative optimization of nutritional status and physical activity to enhance protein metabolism and prevent deconditioning and sarcopenia. In emergency preoperative settings, risk stratification of functional and nutritional status will be important to identify patients with prohibitive risks who may require modification of the type of surgery offered. Intensive perioperative recognition and treatment of hyperglycemia and aggressive perioperative nutritional support may also favorably affect the adverse outcomes.

These inferences must be considered in light of several limitations of the present study, which is based on a cohort of only 89 patients. Clearly, a major weakness of our study is the lack of a control group, which enables direct comparison of the study group in terms of RRs and odds ratios. With use of a control group, this study may have yielded different results. Furthermore, performance of a retrospective review is limited by reliance on the accuracy of documentation by caregivers. Investigator bias looms large in any retrospective study design in which medical records are examined long after hospital discharge. However, consistency of these findings with those in the literature suggests that these confounders exerted minimal influence on the results.

Another limitation was that the determination of poor preoperative functional status and nutritional status was not rigorous. We did not use any standard assessment tools because paucity and inconsistency of the nurse and physician notes made it difficult to use a standard checklist of clinical variables. Instead, we looked for keywords in the medical records that suggested clear evidence of functional and nutritional deficits. It is possible that we may have underestimated the true incidence of preoperative functional deficits and malnutrition.

Retrospective analysis has been shown to underestimate the incidence of true delirium, with physicians and nurses missing or neglecting to document 29% of cases of postoperative delirium when medical records were compared against prospective clinical interviews.28 This finding suggests that the incidence of delirium in this study might have been a conservative estimate.

The accuracy of our definition for delirium may also have been inadequate and may represent more of an assessment of the presence of altered mental status rather than the strict DSM-IV-TR criteria.16 However, a prospective study2 has demonstrated that geriatric patients with a partial diagnosis of delirium closely resemble those with delirium defined according to the DSM-III, with consistent outcomes of a low rate of resolution after discharge and a prolonged hospital stay.

Some may question the choice of the high-risk inclusion criteria based on the prediction model developed by Inouye and colleagues on a cohort of medical rather than surgical patients. We chose this model because of its simplicity and high predictability of delirium among its high-risk subset. The high incidence (60%) of delirium among our high-risk patients fell within the range of other prediction models derived from patients undergoing surgery. This finding suggested that the choice of inclusion criteria did not significantly affect the study results.

Other issues are open for debate but are likely inconsequential to the study conclusion. These issues include whether or not the SUN:Cr ratio of 18 or higher truly reflected dehydration and hypovolemia. Inadequate glycemic control was defined as a serum glucose level of more than 110 mg/dL in the study by van den Berghe et al,23 but we chose a higher level of greater than 150 mg/dL in the present study. Furthermore, not only is hypoalbuminemia a marker of long-term nutritional status but it also fluctuates during acute illness and physiologic stress; it may not be interpreted simply as an indicator for nutritional state.

Also open to criticism may be the criteria used to define suboptimal care. Failure to document either recognition or treatment of various clinical problems (eg, urine output or pain score) has not proved to be associated with adverse outcomes, and designation of suboptimal care may not have been fair. However, the data summarized in Table 4 may be construed as descriptive, and as such, the findings remain striking and noteworthy.

Despite the importance of perioperative nutrition such as enteral feeding, delay or neglect of nutritional support was seen in nearly a third of the patients with poor nutritional status. Inadequate glycemic control was unrecognized or undertreated in 56% of patients with glucose levels greater than 150 mg/dL, despite the knowledge of level I evidence of its correlation with high morbidity and mortality.23 Inadequate postoperative fluid hydration was suboptimally managed in only 22% of the affected patients. One may postulate that the surgical team in a university-affiliated teaching hospital is more educated, and therefore more aggressive, in the treatment of postoperative hypovolemia than hyperglycemia or nutritional deficit. Clearly, opportunities exist for continuing education, improvement of care, and reduction of adverse outcomes when caring for high-risk geriatric patients undergoing major abdominal surgery.

Correspondence: K. Francis Lee, MD, MPH, Department of Surgery, Baystate Medical Center, 759 Chestnut St, Springfield, MA 01199 (leek@hampdencountyphysicians.com).

Accepted for Publication: April 23, 2006.

Author Contributions: Drs Ganai and Lee had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Ganai, K. F. Lee, Merrill, Bellantonio, Brennan, and Lindenauer. Acquisition of data: Ganai, K. F. Lee, Merrill, and M. H. Lee. Analysis and interpretation of data: Ganai, K. F. Lee, Merrill, M. H. Lee, and Brennan. Drafting of the manuscript: Ganai, K. F. Lee, and Merrill. Critical revision of the manuscript for important intellectual content: Ganai, K. F. Lee, M. H. Lee, Bellantonio, Brennan, and Lindenauer. Statistical analysis: Ganai. Administrative, technical, and material support: K. F. Lee, M. H. Lee, and Lindenauer. Study supervision: K. F. Lee, Brennan, and Lindenauer.

Financial Disclosure: None reported.

Francis  JMartin  DKapoor  WN A prospective study of delirium in hospitalized elderly. JAMA 1990;263 (8) 1097- 1101
PubMed Link to Article
Levkoff  SEEvans  DALiptzin  B  et al.  Delirium: the occurrence and persistence of symptoms among elderly hospitalized patients. Arch Intern Med 1992;152 (2) 334- 340
PubMed Link to Article
Dyer  CBAshton  CMTeasdale  TA Postoperative delirium: a review of 80 primary data-collection studies. Arch Intern Med 1995;155 (5) 461- 465
PubMed Link to Article
Marcantonio  ERGoldman  LMangione  CM  et al.  A clinical prediction rule for delirium after elective noncardiac surgery. JAMA 1994;271 (2) 134- 139
PubMed Link to Article
Rolfson  DBMcElhaney  JERockwood  K  et al.  Incidence and risk factors for delirium and other adverse outcomes in older adults after coronary artery bypass graft surgery. Can J Cardiol 1999;15 (7) 771- 776
PubMed
Galanakis  PBickel  HGradinger  RVon Gumppenberg  SForstl  H Acute confusional state in the elderly following hip surgery: incidence, risk factors and complications. Int J Geriatr Psychiatry 2001;16 (4) 349- 355
PubMed Link to Article
Böhner  HHummel  TCHabel  U  et al.  Predicting delirium after vascular surgery: a model based on pre- and intraoperative data. Ann Surg 2003;238 (1) 149- 156
PubMed
Rizzo  JABogardus  STLeo-Summers  LWilliams  CSAcampora  DInouye  SK Multicomponent targeted intervention to prevent delirium in hospitalized older patients: what is the economic value? Med Care 2001;39 (7) 740- 752
PubMed Link to Article
Etzioni  DALiu  JHMaggard  MAKo  CY The aging population and its impact on the surgery workforce. Ann Surg 2003;238 (2) 170- 177
PubMed
Inouye  SKViscoli  CMHorwitz  RIHurst  LDTinetti  ME A predictive model for delirium in hospitalized elderly medical patients based on admission characteristics. Ann Intern Med 1993;119 (6) 474- 481
PubMed Link to Article
Parikh  SSChung  F Postoperative delirium in the elderly. Anesth Analg 1995;80 (6) 1223- 1232
PubMed
Marcantonio  ERGoldman  LOrav  EJCook  FLee  TH The association of intraoperative factors with the development of postoperative delirium. Am J Med 1998;105 (5) 380- 384
PubMed Link to Article
Schneider  FBohner  HHabel  U  et al.  Risk factors for postoperative delirium in vascular surgery. Gen Hosp Psychiatry 2002;24 (1) 28- 34
PubMed Link to Article
Beliveau  MMMultach  M Perioperative care for the elderly patient. Med Clin North Am 2003;87 (1) 273- 289
PubMed Link to Article
Katz  SFord  ABMoskowitz  RWJackson  BAJaffe  MW Studies of illness in the aged: the index of ADL. JAMA 1963;185 (12) 914- 919
PubMed Link to Article
 American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 4th ed, text revision Washington, DC American Psychiatric Association2000;
 Notes on Newton Rhapson method. http://uhaweb.hartford.edu/nagurney/newton.htm. Accessed August 30, 2007
Lynch  EPLazor  MAGellis  JE  et al.  The impact of postoperative pain on the development of postoperative delirium. Anesth Analg 1998;86 (4) 781- 785
PubMed
Kaneko  TTakahashi  SNaka  TInoue  YKaibara  N Postoperative delirium following gastrointestinal surgery in elderly patients. Surg Today 1997;27 (2) 107- 111
PubMed Link to Article
Hamel  MBHenderson  WGKhuri  SFDaley  J Surgical outcomes for patients aged 80 and older. J Am Geriatr Soc 2005;53 (3) 424- 429
PubMed Link to Article
Lee  KF Patient preference- and outcome-based surgical care among octogenarians and nonagenarians. J Am Coll Surg 2006;202 (2) 356- 372
PubMed Link to Article
Aizawa  KKanai  TSaikawa  Y  et al.  A novel approach to the prevention of postoperative delirium in the elderly after gastrointestinal surgery. Surg Today 2002;32 (4) 310- 314
PubMed Link to Article
van den Berghe  GWouters  PWeekers  F  et al.  Intensive insulin therapy in the critically ill patients. N Engl J Med 2001;345 (19) 1359- 1367
PubMed Link to Article
Khuri  SFDaley  JHenderson  W  et al.  Risk adjustment of the postoperative mortality rate for the comparative assessment of the quality of surgical care. J Am Coll Surg 1997;185 (4) 315- 327
PubMed
Inouye  SKBogardus  STCharpentier  PA  et al.  A multicomponent intervention to prevent delirium in hospitalized older patients. N Engl J Med 1999;340 (9) 669- 676
PubMed Link to Article
Cole  MGPrimeau  FMcCusker  J Effectiveness of interventions to prevent delirium in hospitalized patients: a systemic review. CMAJ 1996;155 (9) 1263- 1268
PubMed
Britton  ARussell  R Multidisciplinary team interventions for delirium in patients with chronic cognitive impairment. Cochrane Database Syst Rev 2004;2CD000395
PubMed
Gustafson  YBrännström  BNorberg  ABucht  GWinblad  B Underdiagnosis and poor documentation of acute confusional states in elderly hip fracture patients. J Am Geriatr Soc 1991;39 (8) 760- 776
PubMed

Figures

Place holder to copy figure label and caption
Figure.

Admission source of all 89 geriatric patients undergoing abdominal surgery and discharge destination of the 71 survivors.

Graphic Jump Location

Tables

Table Graphic Jump LocationTable 1. Baseline Characteristics of 89 Patients at High Risk for Delirium
Table Graphic Jump LocationTable 2. Indications for Abdominal Surgery for 89 Patients at High Risk for Delirium
Table Graphic Jump LocationTable 3. Morbidity by Organ System Among 89 Patients at High Risk for Delirium
Table Graphic Jump LocationTable 4. Postoperative Clinical Factors and Incidence of Suboptimal Care Among 89 Patients at High Risk for Delirium
Table Graphic Jump LocationTable 5. Univariate Analyses of Association of Clinical Factors With Adverse Outcomes
Table Graphic Jump LocationTable 6. Adjusted ORs for Independent Predictors of Adverse Outcomes Using Logistic Regression

References

Francis  JMartin  DKapoor  WN A prospective study of delirium in hospitalized elderly. JAMA 1990;263 (8) 1097- 1101
PubMed Link to Article
Levkoff  SEEvans  DALiptzin  B  et al.  Delirium: the occurrence and persistence of symptoms among elderly hospitalized patients. Arch Intern Med 1992;152 (2) 334- 340
PubMed Link to Article
Dyer  CBAshton  CMTeasdale  TA Postoperative delirium: a review of 80 primary data-collection studies. Arch Intern Med 1995;155 (5) 461- 465
PubMed Link to Article
Marcantonio  ERGoldman  LMangione  CM  et al.  A clinical prediction rule for delirium after elective noncardiac surgery. JAMA 1994;271 (2) 134- 139
PubMed Link to Article
Rolfson  DBMcElhaney  JERockwood  K  et al.  Incidence and risk factors for delirium and other adverse outcomes in older adults after coronary artery bypass graft surgery. Can J Cardiol 1999;15 (7) 771- 776
PubMed
Galanakis  PBickel  HGradinger  RVon Gumppenberg  SForstl  H Acute confusional state in the elderly following hip surgery: incidence, risk factors and complications. Int J Geriatr Psychiatry 2001;16 (4) 349- 355
PubMed Link to Article
Böhner  HHummel  TCHabel  U  et al.  Predicting delirium after vascular surgery: a model based on pre- and intraoperative data. Ann Surg 2003;238 (1) 149- 156
PubMed
Rizzo  JABogardus  STLeo-Summers  LWilliams  CSAcampora  DInouye  SK Multicomponent targeted intervention to prevent delirium in hospitalized older patients: what is the economic value? Med Care 2001;39 (7) 740- 752
PubMed Link to Article
Etzioni  DALiu  JHMaggard  MAKo  CY The aging population and its impact on the surgery workforce. Ann Surg 2003;238 (2) 170- 177
PubMed
Inouye  SKViscoli  CMHorwitz  RIHurst  LDTinetti  ME A predictive model for delirium in hospitalized elderly medical patients based on admission characteristics. Ann Intern Med 1993;119 (6) 474- 481
PubMed Link to Article
Parikh  SSChung  F Postoperative delirium in the elderly. Anesth Analg 1995;80 (6) 1223- 1232
PubMed
Marcantonio  ERGoldman  LOrav  EJCook  FLee  TH The association of intraoperative factors with the development of postoperative delirium. Am J Med 1998;105 (5) 380- 384
PubMed Link to Article
Schneider  FBohner  HHabel  U  et al.  Risk factors for postoperative delirium in vascular surgery. Gen Hosp Psychiatry 2002;24 (1) 28- 34
PubMed Link to Article
Beliveau  MMMultach  M Perioperative care for the elderly patient. Med Clin North Am 2003;87 (1) 273- 289
PubMed Link to Article
Katz  SFord  ABMoskowitz  RWJackson  BAJaffe  MW Studies of illness in the aged: the index of ADL. JAMA 1963;185 (12) 914- 919
PubMed Link to Article
 American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 4th ed, text revision Washington, DC American Psychiatric Association2000;
 Notes on Newton Rhapson method. http://uhaweb.hartford.edu/nagurney/newton.htm. Accessed August 30, 2007
Lynch  EPLazor  MAGellis  JE  et al.  The impact of postoperative pain on the development of postoperative delirium. Anesth Analg 1998;86 (4) 781- 785
PubMed
Kaneko  TTakahashi  SNaka  TInoue  YKaibara  N Postoperative delirium following gastrointestinal surgery in elderly patients. Surg Today 1997;27 (2) 107- 111
PubMed Link to Article
Hamel  MBHenderson  WGKhuri  SFDaley  J Surgical outcomes for patients aged 80 and older. J Am Geriatr Soc 2005;53 (3) 424- 429
PubMed Link to Article
Lee  KF Patient preference- and outcome-based surgical care among octogenarians and nonagenarians. J Am Coll Surg 2006;202 (2) 356- 372
PubMed Link to Article
Aizawa  KKanai  TSaikawa  Y  et al.  A novel approach to the prevention of postoperative delirium in the elderly after gastrointestinal surgery. Surg Today 2002;32 (4) 310- 314
PubMed Link to Article
van den Berghe  GWouters  PWeekers  F  et al.  Intensive insulin therapy in the critically ill patients. N Engl J Med 2001;345 (19) 1359- 1367
PubMed Link to Article
Khuri  SFDaley  JHenderson  W  et al.  Risk adjustment of the postoperative mortality rate for the comparative assessment of the quality of surgical care. J Am Coll Surg 1997;185 (4) 315- 327
PubMed
Inouye  SKBogardus  STCharpentier  PA  et al.  A multicomponent intervention to prevent delirium in hospitalized older patients. N Engl J Med 1999;340 (9) 669- 676
PubMed Link to Article
Cole  MGPrimeau  FMcCusker  J Effectiveness of interventions to prevent delirium in hospitalized patients: a systemic review. CMAJ 1996;155 (9) 1263- 1268
PubMed
Britton  ARussell  R Multidisciplinary team interventions for delirium in patients with chronic cognitive impairment. Cochrane Database Syst Rev 2004;2CD000395
PubMed
Gustafson  YBrännström  BNorberg  ABucht  GWinblad  B Underdiagnosis and poor documentation of acute confusional states in elderly hip fracture patients. J Am Geriatr Soc 1991;39 (8) 760- 776
PubMed

Correspondence

CME
Also Meets CME requirements for:
Browse CME for all U.S. States
Accreditation Information
The American Medical Association is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. The AMA designates this journal-based CME activity for a maximum of 1 AMA PRA Category 1 CreditTM per course. Physicians should claim only the credit commensurate with the extent of their participation in the activity. Physicians who complete the CME course and score at least 80% correct on the quiz are eligible for AMA PRA Category 1 CreditTM.
Note: You must get at least of the answers correct to pass this quiz.
Your answers have been saved for later.
You have not filled in all the answers to complete this quiz
The following questions were not answered:
Sorry, you have unsuccessfully completed this CME quiz with a score of
The following questions were not answered correctly:
Commitment to Change (optional):
Indicate what change(s) you will implement in your practice, if any, based on this CME course.
Your quiz results:
The filled radio buttons indicate your responses. The preferred responses are highlighted
For CME Course: A Proposed Model for Initial Assessment and Management of Acute Heart Failure Syndromes
Indicate what changes(s) you will implement in your practice, if any, based on this CME course.
Submit a Comment

Multimedia

Some tools below are only available to our subscribers or users with an online account.

Web of Science® Times Cited: 26

Related Content

Customize your page view by dragging & repositioning the boxes below.

Articles Related By Topic
Related Collections
JAMAevidence.com

The Rational Clinical Examination
Make the Diagnosis: Delirium

The Rational Clinical Examination
Original Article: Does This Patient Have Delirium? Value of Bedside Instruments