With the aging of the US population, issues related to surgical care of older patients have considerable public health significance. The elderly age group requires health care resources at a per capita rate that is higher than for any other age group. Perioperative care of older patients generates a variety of challenges. Age-related disease and declining physiologic functional reserve limit the ability of elderly surgical patients to withstand the stresses of anesthesia, surgery, and recovery. Elderly surgical patients have more perioperative complications than younger patients. 1 Complicating the issue is the difficulty of establishing the functional reserve of the older patients because of deconditioning, age-related disease, and cognitive impairment. Even determining who is an appropriate surgical candidate can be difficult. Further, the paucity of well-designed, definitive research concerning elderly surgical patients has provided clinicians with little guidance.
The chapter on geriatric anesthesia in New Frontiers in Geriatrics Research noted that geriatrics issues affect all areas of anesthesiology: preoperative evaluations, intraoperative management, postoperative care, and chronic pain management. 2 The chapter began with a review of age-related physiologic changes in the nervous, cardiovascular, respiratory, and urinary systems (see New Frontiers, pp. 10–14). In addition, it provided a synopsis of alterations in pharmacokinetics and pharmacodynamics seen in the elderly patient (see p. 14). This information provided a framework for the discussion of existing research and the development of recommendations for needed research. The result was 41 agenda items pertinent to anesthesiology, management of pain, and the perioperative care of the elderly patient. In addition, three issues were defined as Key Questions:
Anes KQ1: What preoperative assessments are useful in developing patient management plans for surgeries common in the elderly population?
Anes KQ2: Can proper choice of postoperative analgesic techniques reduce postoperative morbidity or improve functional status at discharge?
Anes KQ3: How can postoperative pulmonary complications in the elderly patient be reduced?
The Key Questions specifically address the pre- and postoperative care of the older patient. The literature reviewed for New Frontiers suggested that there is no substantive difference in outcome between general anesthesia and regional anesthesia. It has been argued that this is the most significant choice anesthesiologists can make regarding intraoperative care. Thus, investigations regarding other intraoperative management issues were thought likely to be of low yield. Given limited research funding, the decision was made to focus the research agenda on areas more likely to be fruitful.
This review updates the literature review of New Frontiers and identifies progress since the initial review of 2004. It also addresses emerging areas of research that warrant further investigation with regard to the perioperative care of the elderly person. New Frontiers laid out a broad agenda for research into issues of perioperative care of elderly surgical patients. That several of the questions have yet to be answered is not discouraging. To be answered satisfactorily, many of the issues require large multicenter clinical trials. Several years are required to conceive and design such trials, obtain funding, collect and analyze data, and publish the results. Thus, the number of publications addressing items in the research agenda put forward in New Frontiers is encouraging. It indicates that researchers already are aware of areas requiring more investigation.
This update identifies two emerging areas that require further study: perioperative glucose control and the role of intra- and postoperative anesthetic and analgesic management, specifically, regional versus general anesthesia. Two new agenda items have therefore been added to the research agenda in geriatric anesthesiology. (See the section New Horizons in Geriatric Anesthesia at the end of the chapter.)
As more data regarding these and issues previously identified by the project become available, information can be incorporated into clinical pathways aimed at improving medical and functional outcomes in elderly surgical patients. Anesthesiologists are involved throughout the continuum of care for every older surgical patient and will have a key role in investigating and implementing interventions to ensure that older patients have optimal surgical outcomes.
A literature search was performed on the National Library of Medicine’s PubMed database in July 2005. Anesthesiology and intraoperative care were searched as major MeSH headings and combined with various MeSH terms denoting advanced age: age, geriatric assessment, aged, aged 80 and over, frail elderly, longevity, and geriatrics. The following limits were placed on all searches: aged: 65+ years, the English language, humans, and the years 2000–2005. The anesthesiology portion of the search yielded 62 articles, and the intraoperative care segment resulted in 200 articles.
A separate search of the PubMed database was performed in June 2005. In this search, the topic anesthesiology was combined with the following terms for advanced age: 65 or older, aged, and geriatric. Limits placed on the search included: the English language, core clinical journals, and the years 2000–2005. This search located 107 articles.
Additionally, reference lists from review articles included in the search results were examined for relevant articles.
Finally, separate searches were performed for each individual agenda item, using appropriate search terms and limiting the search as described.
All articles obtained in the various search strategies were reviewed to determine their relevance to the research agenda set forth in the 41 agenda items and three Key Questions in the anesthesiology chapter of New Frontiers. Papers that are relevant to the research agenda are briefly summarized herein, and recommendations are made as to whether any original agenda items need to be altered or omitted in light of recent findings.
Anes 1 (Level B): Prospective epidemiologic studies are needed to describe the relative frequency of various outcomes characteristic of older surgical patients for the most common types of surgery.
New Research Addressing This Question: No research published since the initial literature review was found.
Modification of This Question in Light of New Research: This question remains vital. Descriptions of the outcomes of older surgical patients will be the cornerstone for the design, testing, and implementation of strategies for appropriate preoperative evaluation of and intervention for aged patients (see Anes 3).
Anes 2 (Level B): Once better understanding of characteristic outcomes of specific types of surgery for older patients is attained, patient- and surgery-specific risk factors for geriatric complications should be identified by multivariate analysis that would stratify surgical risk as low, intermediate, or high, depending on type of surgery.
New Research Addressing This Question: No research published since the initial literature review was found.
Modification of This Question in Light of New Research: Anes 2 should remain on the research agenda. Information obtained from these types of studies will be invaluable in the design, testing, and implementation of preoperative testing and interventional strategies for improving outcome in geriatric surgical patients.
Anes 3 (Level B): The positive predictive value of preoperative assessment instruments should be determined in prospective nonrandomized or prospective cohort trials.
New Research Addressing This Question: The ongoing growth of the elderly age group has serious implications for preoperative evaluation clinics. Elderly patients are the most frequent users of surgical services and often require more extensive preoperative evaluation than younger patients having equivalent procedures. Thus, the challenge is to evaluate increasing numbers of patients needing relatively thorough examinations in a manner that is both cost-efficient and able to correctly identify conditions that may influence perioperative management. The use of questionnaires is one method by which this may be achieved. Ideally, these instruments could differentiate between patients whose medical status dictates a detailed preoperative medical evaluation and those who require a less extensive evaluation. Patients identified as needing an extensive preoperative work-up could be referred directly to a physician. Others could be sent first to a midlevel provider and referred to a physician only if pertinent issues were identified.
Hilditch et al developed a screening questionnaire to determine which patients need to be evaluated by an anesthesiologist before the day of surgery. The 17-question instrument contained closed-end questions regarding general health, exercise tolerance, family history, and previous anesthetic history. Patient responses to the items in the questionnaire were compared with the history obtained by an anesthesiologist. All questions except those regarding kidney disease and previous personal and family anesthetic history had at least good validity as determined by κ-coefficients or percentage agreement with the anesthesiologist’s assessment, or both. 3
Another approach to preoperative evaluation is to have allied health personnel screen patients to determine those who need to be examined by an anesthesiologist before going to surgery. A study by van Klei et al assessed the accuracy of specially trained nurses in preoperative health assessment. In this study, experienced postanesthesia care unit nurses received training in medical history taking and physical examination and attended lectures on a variety of medical subjects related to perioperative care, including those concerning the elderly patient. Patients (N = 4540) were first assessed by a trained nurse and then by an anesthesiologist. The nurses were asked to determine whether a patient was ready for surgery or required further evaluation. The nurse’s assessment was compared with that of the anesthesiologist. Nurses were able to identify patients who were not ready for surgery with a sensitivity of 83% (95% confidence interval [CI]: 79% to 87%), a specificity of 87% (86% to 88%), a positive predictive value of 34% (31% to 38%), and a negative predictive value of 98% (98% to 99%). Therefore, although nurses often incorrectly identified patients as being not ready for surgery and referred them for further evaluation when none was needed, they rarely incorrectly classified patients requiring more extensive work-up as being ready for surgery. Although there was some inefficiency to this approach (anesthesiologists still needed to see patients who needed minimal assessment), it nonetheless provided for more efficient resource utilization than the traditional model by eliminating the need for a physician-conducted preoperative evaluation in certain patients. Most importantly, patients who required physician-conducted preoperative examinations rarely were misidentified. 4
Given the importance of malnutrition in predicting adverse surgical outcomes, 5 it is vital to preoperatively identify malnourished patients, since this information may influence surgical decision making. 6 The Mini Nutritional Assessment (MNA) tool can be used to evaluate nutritional status during a preoperative evaluation. 7–9 However, because it requires anthropomorphic measurements and the completion of a questionnaire, it is time-consuming and may not be appropriate in the setting of a focused preoperative examination. The Mini Nutritional Assessment–Short Form (MNA–SF) consists of six items extracted from the original MNA: five questions and body mass index. 10 The MNA–SF can be administered quickly and has a sensitivity and specificity of 100% and 69.5%, respectively, for determining the absence of overt malnutrition in elderly surgical patients. Although the positive predictive value is only 19.4%, the negative predictive value was 100% (if the MNA–SF is negative, the patient does not have overt malnutrition). 11
Modification of This Question in Light of New Research: The MNA–SF successfully identifies patients with overt malnutrition. Given that this condition may alter surgical planning, 6 it would be reasonable to evaluate the MNA–SF in a prospective trial to determine whether its application improves perioperative outcomes in older surgical patients (see Anes 4).
The Hilditch questionnaire and nurse evaluation studies are encouraging. However, neither approach focused on the elderly age group. Given the complexity of older surgical patients, these instruments need to be validated in elderly cohorts. Further, taken together, the results of the two studies suggest that a third approach may be viable: the use of a questionnaire to triage patients requiring extensive evaluation directly to a physician. Others would be seen by a midlevel provider and referred to a physician only if relevant issues were identified. Therefore, Anes 3 should neither be modified nor eliminated.
Anes 4 (Level A): Following evaluation of preoperative assessment instruments (Anes 3), prospective randomized trials should be performed to determine whether the application of these metrics could improve outcomes for elderly surgical patients by altering perioperative intervention, surgical timing, the type or extent of surgery, or postoperative management.
New Research Addressing This Question: No research published since the initial literature review was found.
Modification of This Question in Light of New Research: Development of instruments described in Anes 3 will be beneficial only if they result in interventions to improve outcomes. Therefore, Anes 4 should remain unchanged.
Anes 5 (Level B): Prospective cohort studies are needed to determine whether assessment of the older surgical patient’s preoperative functional status affects surgical decision making or perioperative care.
New Research Addressing This Question: A prospective cohort study of 544 patients examined the relative contributions of preoperative health status and intraoperative factors on the incidence of complications in older surgical patients. Multivariate analysis demonstrated that decreased functional status (defined as ≥ 3, where 1 = no limitation of activities, 2 = intermittent limitation, 3 = mild limitation, 4 = moderate limitation, and 5 = severe limitation) is an independent risk factor for adverse postoperative neurologic events. 12 In a prospective cohort study of elderly patients having hip and knee arthroplasty, even subtle declines in preoperative functional status was found to predict postoperative delirium. 13 However, our search strategy uncovered no studies addressing whether baseline functional status affects surgical decision making or perioperative care. No specific recommendations regarding functional status of the heart, pulmonary system, kidneys, mobility, or cognitive status are yet available.
Modification of This Question in Light of New Research: No modification of Anes 5 is recommended. Altering surgical care in light of predictors of adverse outcome is a reasonable approach to the care of older surgical patients. The advent of newer, less invasive alternatives to traditional open procedures offers options where none were previously available. 14 Assessing whether preoperative functional status leads to alterations in perioperative care lays a foundation for the studies proposed in Anes 6. In addition, studies proposed in Anes 5 will provide a measure of the value that those involved in the surgical decision-making process and providing perioperative care place on functional assessment. If the predictive value of functional status on perioperative outcomes is shown not to be appreciated, there will be opportunities to improve care through educational efforts emphasizing its importance.
Anes 6 (Level A): Depending on findings of prospective cohort studies (Anes 5), randomized trials should be performed to determine whether preoperative functional status assessment of elderly patients changes decisions about type or timing of surgery, or pre- or postoperative care strategies and outcomes.
New Research Addressing This Question: No research published since the initial literature review was found.
Modification of This Question in Light of New Research: In an era of limited health care resources, studies such as those suggested by Anes 6 are vitally important. As more options for less invasive surgical procedures evolve (see Anes 5), selecting the correct procedure for a given patient will become more complicated. A key public health issue will be to select procedures in order to maximize improvement in functional status while minimizing the risk of complications and resource utilization. For example, is there a level of preoperative functional status at which the risk of complications for a given orthopedic procedure is sufficiently high such that it would better to perform a less invasive or temporizing procedure? Conversely, is there a subgroup of patients for whom more heroic operations are indicated? The goal of these studies would be to quantify this information for various elderly surgical populations and develop metrics that would enable surgical decision making to relate more directly to current and predicted functional status.
Anes 7 (Level B): Cross-sectional or prospective cohort studies are needed to determine by multivariate analysis whether there is an association between pre-existing cognitive impairment, depression, or alcohol abuse and adverse outcomes in geriatric patients.
New Research Addressing This Question: It is well known that significant pre-existing cognitive impairment, depression, or alcohol abuse predict delirium in elderly medical patients. 15,16 Alcohol abuse and significant cognitive impairment also predict postoperative delirium in elderly noncardiac surgical patients. 17 However, these issues have not been extensively studied in elderly surgical patients. This is important because, unlike medical patients, older surgical patients are generally not acutely ill at the time of surgery. Also, the physiologic stresses imposed by anesthesia and surgery differ from those of medical illness. Thus, studies in this area are essential.
A prospective cohort study examining predictors of postoperative delirium in 427 elderly total hip and total knee arthroplasty patients demonstrated that even subtle preoperative cognitive impairment predicts postoperative delirium. Forty six (10.8%) developed postoperative delirium. Preoperative verbal intelligence as measured by the American National Adult Reading Test 18 and gross cognitive status as measured by the Mini-Mental State Examination (MMSE) 19 were similar between those who did and did not develop postoperative delirium. However, as a group, patients who subsequently developed postoperative delirium were found to have decreased verbal fluency (as measured by the Controlled Oral Word Association Test 20), verbal learning (as measured by the Auditory Verbal Learning Test 21), and concentration (as measured by the Stroop Color-Word Test 22). 13 Gruber-Baldini et al performed a multicenter longitudinal cohort study in 674 elderly hip fracture patients that examined the outcomes associated with cognitive impairment. This group found that that prefracture or in-hospital cognitive impairment as indicated by an MMSE score lower than 24 or the presence of delirium predicted poor functional outcome at 2 and 12 months postoperatively. In addition, incident cognitive dysfunction predicted later functional and social impairment. 23
Modification of This Question in Light of New Research: Although these studies provide some guidance with regard to cognitive impairment, further work needs to be done to make sure that the results are applicable across a wide range of surgical procedures. In addition, further research is required to determine which tests are most useful for the preoperative evaluation of cognitive function (see Anes 33). Finally, given the prevalence of depression and alcohol abuse in the elderly age group,>). Finally, given the prevalence of depression and alcohol abuse in the elderly age group, it is vital to examine the role they play in postoperative outcomes. 24–30
Anes 8 (Level A): For any association that is established by cross-sectional cohort studies between cognitive impairment, depression, or alcohol abuse and adverse outcomes in elderly surgical patients (Anes 7), prospective randomized trials should be performed to determine the effect of pre- or postoperative interventions on these adverse outcomes.
New Research Addressing This Question: No research published since the initial literature review was found.
Modification of This Question in Light of New Research: No modification of Anes 8 is recommended.
Anes 9 (Level B): Cross-sectional or retrospective case-control studies are needed to identify the incidence of adverse cardiac or thrombotic-embolic complications in elderly patients undergoing surgery with and without preoperative β-blockade, antithrombotic or antiplatelet therapy, or with a hematocrit above a target value. These studies should be in surgeries identified as having an intermediate or high risk for related complications.
New Research Addressing This Question: A large multicenter prospective observational study addressed the effects of perioperative aspirin for patients undergoing coronary artery bypass grafting. Aspirin (650 mg) administered within 48 hours after revascularization was found to be independently associated with a lower incidence of death, nonfatal myocardial infarction, stroke, renal failure, and bowel infarction. Multivariate analysis did not reveal any other factors to be associated with the reductions of these adverse outcomes. In addition, discontinuation of aspirin use before surgery was found to be an independent risk factor for mortality (odds ratio [OR] = 1.79, 95% CI: 1.18 to 2.69). The risk of hemorrhage, gastritis, infection, and impaired wound healing was not increased with aspirin use. 31
No studies addressing the issue of perioperative aspirin therapy in a noncardiac surgical population were identified.
A number of previously published prospective studies demonstrate an association between perioperative β-blockade and improved cardiac outcomes, particularly in high-risk patients undergoing high-risk procedures. 32–35 However, in a recent large retrospective study involving nearly 800,000 patients, perioperative β-adrenergic blockade was found to be associated with an increase in in-hospital mortality in patients who had a low a priori risk of cardiac morbidity and mortality (OR = 1.43, 95% CI: 1.29 to 1.58) (see Figure 2.1). 36 Ongoing studies will refine understanding of the role of perioperative β-blockade in elderly surgical patients. 37–39
To receive a copy of Figure 2.1, please contact Ellen Baumritter at
ebaumritter@americangeriatrics.org
Postoperative thromboembolic complications are common and an area of active study (see Anes 10). A retrospective case-control study demonstrated that enoxaparin is superior to warfarin monotherapy for early venous thromboprophylaxis following hip and knee arthroplasty in postmenopausal women (OR 8.6, P < .0001 for any thrombosis, OR 11.3, P < .0001 for proximal thrombosis). 40 However, in another retrospective case-control study, thomboprophylaxis with tinzaparin was found to be associated with an increased incidence of serious postoperative bleeding. 41 Thus, although low-molecular-weight heparins may be more effective than warfarin therapy in preventing venous thromboembolism, they carry the risk of hemorrhagic complications. Multimodal thromboprohylaxis may offer a way to minimize thrombotic complications and bleeding associated with prophylactic measures. In a large retrospective case series, extensive multimodal thromboprophylaxis (preoperative discontinuation of procoagulant medications, autologous blood donation, hypotensive epidural anesthesia, intravenous heparin during surgery and before femoral preparation, aspiration of intramedullary contents, pneumatic compression, thigh-high stockings, early mobilization, and chemoprophylaxis with aspirin or warfarin for 4 to 6 weeks postoperatively) decreased the incidence of thrombosis and pulmonary embolism in hip arthroplasty patients in comparison with historical controls. 42
No studies were identified regarding hematocrit value.
Modification of This Question in Light of New Research: There is little question that in high-risk patients undergoing high-risk surgery, perioperative β-blockade reduces adverse cardiac outcomes. Extant studies will help determine its role in patients with lower risk. 37–39
Perioperative antiplatelet therapy in the form of aspirin has a salutary effect on patients undergoing coronary artery bypass grafting. However, the role of aspirin and other antiplatelet therapy in elderly noncardiac surgical patients still needs to be defined.
Although the efficacy and safety of various antithrombotic strategies have been studied in large prospective randomized trials (see Anes 10), none focused exclusively on elderly patients. This is significant, given the frequency of polypharmacy in older patients and the drug interactions associated with warfarin. 43 Clearly, there is a need for investigations in this area. In addition to retrospective case-control studies, one option for obtaining hypotheses-generating data might be to perform post hoc analyses on data from the elderly patients who participated in the randomized trials. These studies should examine the incidence of adverse reactions as closely as prevention of thrombosis. Subsequent prospective randomized trials could be designed on the basis of these results. Further, there is a need for well-designed retrospective studies examining the utility of multimodal approaches to thromboprophylaxis. This approach is especially attractive for geriatric patients, given that it minimizes pharmacologic intervention.
Alternative medicines are used by a large proportion of the population, and many are associated with excessive bleeding. 44,45 However, to date, no larger studies have associated them with increased perioperative blood loss, transfusion, or bleeding complications. In an era of limited research funding, investigations into the perioperative implications of the use of these substances should have low priority.
A more appropriate use of research resources might be to examine ways to improve administration of known therapies for preventing thrombotic complications. Despite evidence of their efficacy, they are not used universally. For example, although the Netherlands has national guidelines for thromboprophylaxis, compliance is not optimal. 46 Computerized alerts and other systems changes may improve delivery of thromboprophylaxis, 47,48 and there is a need to evaluate the feasibility of broad application of these measures.
Finally, research is still required to determine appropriate target hemoglobin levels in older surgical patients. The American Society of Anesthesiologists Practice Guidelines for Perioperative Blood Transfusion and Adjuvant Therapies states that patients usually should receive transfusions for hemoglobin levels of ≤ 6 g/dL and rarely for levels > 10. 49 However, this guideline does not specifically consider elderly patients. Further, higher hemoglobin levels may be beneficial in the elderly patient. 50 Therefore, there is a critical need for studies in this area.
Anes 10 (Level A): For any association in elderly patients of cardiac thrombotic-embolic complications with a specific preoperative therapy or hematocrit level, prospective cohort or randomized studies are needed to determine if pre- or intraoperative therapies would reduce the complications.
New Research Addressing This Question: A number of recent large prospective randomized trials have examined the efficacy of a variety of thromboprophylactic strategies, including newer agents such as factor Xa and thrombin inhibitors. 51–60 However, as noted (see Anes 9), none focused solely on older patients.
Modification of This Question in Light of New Research: Prospective studies regarding cardiac complications and β-blockers have been completed (see Anes 9). As noted however, questions remain. Ongoing studies may clarify the issue. 37–39
Although there has been active investigation of various thromboprophylactic therapies, these studies have not addressed older patients per se. Therefore, more research is required to assure that these interventions have acceptable risk-benefit ratios in this highly vulnerable population.
Studies regarding hematocrit levels and antiplatelet therapy are still needed.
As a whole, studies suggested by Anes 10 address a broad range of critical issues in elderly surgical patients. Although they will be difficult to design and execute, these studies will provide information that could significantly alter surgical care for older patients.
Anes 11 (Level D): Prospective nonrandomized investigation of the effect of perioperative temperature management on surgical morbidity in the geriatric population is needed. These investigations should be conducted under conditions where either the surgery is physiologically very challenging or the older patient carries a high burden of comorbidity. Cardiac, respiratory, bleeding, and renal outcomes would be the primary focus of these investigations.
New Research Addressing This Question: The search did not locate any articles that directly addressed Anes 11. However, two studies addressed other outcomes related to perioperative temperature management. In one study, 338 mostly elderly patients were randomized to either aggressive forced-air intraoperative warming therapy or standard care during a variety of noncardiac procedures. Patients who were randomized to aggressive intraoperative warming had relative risk reduction of 46% for the development of pressure ulcers. 61
In an interesting study, Yucel et al performed a post hoc analysis of a cohort of patients with colorectal cancer who were randomized to normothermic versus hypothermic (34.5°C) temperature management during primary resection. The groups were demographically similar, including information on tumor grade and stage. 62 With up to 9 years of follow-up, there was no difference in cancer-free or overall survival. However, in the original study, hypothermia was associated with increased incidence of infection, decreased wound collagen deposition, and increases in days to first solid food, suture removal, and hospital discharge. 63 Given the strong evidence of other adverse outcomes associated with it, 63–68 the results of this study should not be viewed as an endorsement of perioperative hypothermia. That said, the authors are to be commended for evaluating long-term outcomes associated with perioperative hypothermia. Across the spectrum of geriatric surgical care, more studies addressing the long-term implications of perioperative management are needed.
Modification of This Question in Light of New Research: As noted, no studies have been published directly addressing Anes 11 since New Frontiers was published. In addition, recent evidence suggests that advanced age is a risk for developing intraoperative hypothermia. 69 Thus, Anes 11 should remain on the research agenda. However, relevant outcomes should be expanded to include surgical site infections. 63
Anes 12 (Level D): Retrospective or cross-sectional research studies should be conducted to identify any relationship between the use or timing of perioperative antibiotic therapy and postoperative pneumonia or wound infection. Differences, if any, between younger and older patients undergoing the same type of surgery should also be examined.
New Research Addressing This Question: A retrospective study examined the effect of intraoperative redosing of prophylactic cefazolin on surgical site infection in cardiac surgery lasting more than 240 minutes after the first dose was given. Most of the 1548 patients were elderly. The overall risk of surgical site infection was similar between those who had intraoperative redosing (9.4%) and those who did not (9.3%) (OR 1.01, 95% CI: 0.70 to 1.47). However, redosing reduced the incidence of surgical site infection in patients whose procedures lasted longer than 400 minutes (7.7% versus 16.0%, OR 0.44, 95% CI: 0.23 to 0.86). 70
Modification of This Question in Light of New Research: This study is an encouraging start. However, its results may not be generalizable to the noncardiac surgical population. Further, no studies have addressed prophylactic antibiotic use and the risk of pneumonia. Likewise, there have been no studies to determine what, if any, differences exist between younger and older patients. Anes 12 should remain on the research agenda unchanged.
Anes 13 (Level C): Depending on the findings in Anes 12, prospective randomized studies should be used to determine whether preoperative or postoperative antibiotic therapies reduce complications related to infections in elderly surgical patients.
New Research Addressing This Question: No research published since the initial literature review was found.
Modification of This Question in Light of New Research: Surgical site infections are a major source of postoperative morbidity and mortality and may be responsible for ~$1.6 billion in excess hospital charges annually in the United States. 71 Determining effective strategies to prevent complications from perioperative infections will have important public health and patient care implications. Anes 13 should not be altered.
Anes 14 (Level D): Multicenter case-control or prospective cohort studies should be performed to determine whether receiving or not receiving blood in the perioperative period affects the incidence of perioperative infection and immunosuppression in elderly patients. Multivariate analysis would be required to separate the effect of homologous blood transfusion from the comorbid conditions making transfusion more likely.
New Research Addressing This Question: The search revealed two articles related to Anes 14. The first was a small retrospective analysis of 56 patients undergoing repair of ruptured abdominal aortic aneurysms. Respiratory complications (prolonged ventilation, infection, or respiratory distress) occurred in 40.0% of patients receiving salvaged auto-logous blood and in 12.5% patients who did not (P = .047). However, the analysis did not consider whether this may have been a function of comorbid conditions. Also, 95% of patients who received salvaged autologous blood had suprarenal aortic cross-clamping, compared with only 63% in the group that did not receive salvaged blood. These, among other reasons, limit the applicability of this study to a general elderly surgical population. 72
The results from the second study may be more broadly applicable. Tang et al performed a single-center prospective cohort study of 2809 patients undergoing elective resection of the colon and rectum to determine the risk factors for surgical site infection. Most patients were older than 56 years. By the use of multivariate analysis, homologous blood transfusion was found to be an independent risk factor for overall surgical infection (OR = 2.0, 95% CI: 1.1 to 3.3 for 1 to 3 units transfused, OR = 6.2, 95% CI: 4.2 to 10.2 for ≥ 4 units transfused). Interestingly, of the 14 risk factors found to be significant via multivariate analysis, only blood transfusion was a significant risk factor for all surgical site infections examined (incisional and organ-space with and without anastomotic leakage). 73
Modification of This Question in Light of New Research: The study by Tang et al is well done. Nevertheless, more research is needed, specifically, studies more closely examining the effects of age and including other populations for whom surgical site infections may be devastating, such as orthopedic and cardiac surgical patients. In addition, the studies must be designed and analyzed with a view toward specific implications for improving practice.
Anes 15 (Level C): If perioperative infection and immunosuppression in older surgical patients are shown to be associated with receiving blood in the preoperative period (Anes 14), alternative strategies such as delaying surgery or erythropoietin therapy should then be compared with blood transfusion in prospective cohort studies, because a randomized trial could not be justified.
New Research Addressing This Question: No research published since the initial literature review was found.
Modification of This Question in Light of New Research: An aging population places strain on health care resources. Therefore, it is essential to examine simple, inexpensive strategies such as delaying surgery to improve outcome. Anes 15 should remain on the research agenda unchanged.
Anes 16 (Level D): Cohort or case-control studies are needed to determine the relationship in older surgical patients between perioperative termination of anticoagulation and thromboembolic or bleeding risk.
New Research Addressing This Question: No research published since the initial literature review was found.
Modification of This Question in Light of New Research: Given the frequency with which older surgical patients require long-term anticoagulation for mechanical heart valves, chronic congestive heart failure, and history of thrombosis, as well as the variability of perioperative anticoagulant bridging therapy, studies of this nature are imperative. Anes 16 should remain on the research agenda unchanged.
Anes 17 (Level C): The effect of timing of termination and resumption as well as the temporizing use of antiplatelet agents in older surgical patients should be compared in case-control or prospective cohort studies.
New Research Addressing This Question: No research published since the initial literature review was found.
Modification of This Question in Light of New Research: Anes 17 should remain on the research agenda unchanged.
Anes 18 (Level B): In prospective cohort studies the incidence of perioperative hypotension, aspiration, and renal insufficiency should be compared in elderly patients undergoing standard nothing-by-mouth orders before surgery and in elderly patients who would be allowed clear liquids closer to the time of surgery. This study would need to be conducted in:
patients undergoing specific types of procedures where liberalization of fluid intake is not contraindicated for surgical reasons,
patients undergoing procedures that place them at greater risk for developing hypovolemia (eg, bowel prep), and
instances where preoperative hypovolemia may contribute to complications (eg, angiographic procedures).
New Research Addressing This Question: No research published since the initial literature review was found.
Modification of This Question in Light of New Research: Cardiac catheterizations are common in the United States, and most are performed on elderly patients. Aging is associated with both chronic dehydration and reduced creatinine clearance, and these two factors are fundamental determinants of contrast nephropathy. 74 Prevention of contrast nephropathy is an area of active investigation, but those studies have not reached specific conclusions or recommendations regarding care of the elderly patient in angiographic procedures. 74
In January 1999 the American Society of Anesthesiologist Task Force on Preoperative Fasting issued new guidelines for preoperative fasting and the use of pharmacologic agents to reduce the risk of pulmonary aspiration. 75 The revised guidelines call for preoperative fasts of 6, 4, and 2 hours for solids, nonclear liquids, and clear liquids, respectively. In theory, the preoperative fast may now have a lesser effect on volume status and may reduce the incidence of perioperative hypotension and renal insufficiency. That said, chronic dehydration is common in the elderly patient; its incidence increases with age, and it is often misdiagnosed. 76–78 Given the frequency of baseline chronic dehydration, it is unlikely that the new preoperative fasting guidelines will have a significant effect on the incidence of hypotension and renal insufficiency. However, there remains need for studies examining this issue. Further, the new guidelines may increase the incidence of aspiration in elderly surgical patients, a group that already is at high risk. 79,80 It is critical that the effect of the new preoperative fasting guideline on the incidence of aspiration be examined.
Anes 19 (Level B): Prospective studies that better identify patient and procedural risk factors for respiratory failure, aspiration, and pneumonia in elderly surgical patients are needed.
New Research Addressing This Question: No research published since the initial literature review was found.
Modification of This Question in Light of New Research: Pulmonary complications are common after surgery. 81,82 Although patient and procedural risk factors have been studied, little is known about them in the elderly age group. Recent awareness of the role of obstructive sleep apnea in postoperative pulmonary complications, along with its prevalence and the unreliability of traditional markers for the syndrome in the elderly patient, 83,84 make studies suggested by Anes 19 even more important. Therefore, it should remain on the research agenda unchanged.
Anes 20 (Level A): Randomized trials are needed to determine if respiratory monitoring or O2 therapy can reduce the incidence of respiratory failure in elderly surgical patients.
New Research Addressing This Question: Though there are ongoing studies addressing this issue, 85 none have been published to date.
Modification of This Question in Light of New Research: The interventions suggested by Anes 20 require little effort or expense. If proved to be effective, they will provide an elegant solution to a significant problem. Therefore, Anes 20 should not be altered or eliminated.
Anes 21 (Level A): Randomized studies of prophylactic antibiotics, changes in pharyngeal instrumentation, or the way feeding is advanced are needed to determine whether practice changes reduce aspiration and postoperative pneumonia in elderly surgical patients. (See also the Key Questions at the beginning of the chapter.)
New Research Addressing This Question: No research published since the initial literature review was found.
Modification of This Question in Light of New Research: Anes 21 should remain on the research agenda unchanged. The trend toward earlier discharge and feeding in abdominal surgery patients makes studies regarding the effects of early advancement of feeding even more pertinent.
Anes 22 (Level B): Cross-sectional studies capable of identifying any relationship in elderly surgical patients between intensive nutritional support in high-risk surgery and functional status on discharge (eg, chronic respiratory failure, ambulation, independent living) are needed.
New Research Addressing This Question: No research published since the initial literature review was found.
Modification of This Question in Light of New Research: Functional status may be the most significant outcome measure for elderly surgical patients. Therefore, studies examining the association between perioperative interventions and functional status are especially important. Anes 22 should remain on the research agenda unchanged.
Anes 23 (Level A): Data from studies of associations between nutritional support and postoperative functional status after high-risk surgery in elderly patients (Anes 22) should be used to design prospective cohort or randomized controlled trials comparing feeding strategies in elderly patients at risk for malnutrition and muscle wasting following major surgery.
New Research Addressing This Question: No research published since the initial literature review was found.
Modification of This Question in Light of New Research: Anes 23 should remain on the research agenda unchanged.
Anes 24 (Level C): Randomized trials of interventions that might attenuate postoperative catabolism or facilitate the transition to anabolism in the elderly patient are needed.
New Research Addressing This Question: No research published since the initial literature review was found.
Modification of This Question in Light of New Research: Anes 24 should remain on the research agenda unchanged.
Anes 25 (Level C): Large cross-sectional studies describing analgesic practice and its complications in the elderly surgical patient are needed.
New Research Addressing This Question: Frasco et al examined the effect of the Pain Initiative of the Joint Commission for Accreditation of Healthcare Organizations (JCAHO) on perioperative analgesic care. 86 The JCAHO Pain Initiative standards recognize that pain is often undertreated in surgical patients and are an attempt to enhance perioperative analgesic care by improving assessment and promoting aggressive treatment of pain. 87 In the Frasco study, 541 patients treated by the use of the JCAHO Pain Initiative guidelines were compared with matched historical controls who had surgery in the period immediately before the JCAHO Pain Initiative was implemented. Mean age (± S.D.) in the treatment group was 63.1 (± 14.3) years and 63.9 (± 13.8) years in the historical controls. However, patients ranged from 15 to 93 years of age. Those whose care was delivered according to the guidelines received more opiates, but also required more treatment for postoperative nausea and vomiting. There was no difference between groups in the incidence of respiratory depression requiring intervention. 86
Modification of This Question in Light of New Research: Anes 25 should remain on the research agenda unchanged. The study described did not focus on elderly surgical patients. Given the alterations in baseline opiate sensitivity 88,89 and decreases in metabolism and elimination capacity in the elderly patient (see New Frontiers, pp. 13–14), the studies suggested in Anes 25 remain vitally important and will form the cornerstone for studies based on Anes 26.
Anes 26 (Level B): Depending in part on findings of large, descriptive studies of analgesia in elderly patients (Anes 25), prospective cohort studies are needed to determine the effect of analgesic modes (patient-controlled versus as-needed versus scheduled dosing), route of administration, the role of nonopioid adjunctive drugs, and nonpharmacologic interventions. These investigations must define a balance between adequate analgesia and reduction of the incidence of adverse drug events in the elderly patient.
New Research Addressing This Question: No research published since the initial literature review was found.
Modification of This Question in Light of New Research: The information provided by studies related to this question will provide important preliminary data for trials outlined in Anes 27. Anes 26 should remain on the research agenda unchanged.
Anes 27 (Level A): Prospective randomized controlled trials comparing outcomes with analgesic programs specific to types of surgery are needed to determine whether analgesic regimes designed for the elderly patient reduce in-hospital morbidity or improve functional status on discharge. (See also the Key Questions at the beginning of the chapter.)
New Research Addressing This Question: In a prospective observational study to determine the risk factors for postoperative delirium in elderly patients undergoing total hip and knee arthroplasty, patients whose postoperative analgesic regimen consisted of continuous perineural infusions of local anesthetics supplemented with small amounts of opiates were found to have a 59% lower incidence of postoperative delirium than those who received patient-controlled analgesia intravenous opiates alone. 90 The study did not address the issue of functional status on discharge.
Modification of This Question in Light of New Research: Though the study described was not a randomized trial, its results suggest that investigations in this area may be fruitful. Future trials should be randomized and consider hospital resource utilization and long-term functional status as well as in-hospital morbidity and functional status on discharge. Further, the utility of multimodal analgesic approaches and novel analgesic adjuncts, such as dexmedetomidine, 91,92 should be examined in trials involving operations that are not amenable to regional anesthetic and analgesic techniques.
Anes 28 (Level D): Improved tools for the assessment of pain in the cognitively impaired elderly patient should be developed.
New Research Addressing This Question: Pain is often inadequately treated in the cognitively impaired elderly patient. 93 It is a subjective experience, and significant cognitive impairment limits patients’ abilities to complete self-report measures. 94,95 Therefore, the development of appropriate measures of pain in elderly patients with cognitive impairment is imperative. The challenge is to develop tools that do not rely on self-report of symptoms.
Several new instruments for the assessment of pain in cognitively impaired older patients have been introduced recently. The first is the DOLOPLUS 2. It consists of 10 items across three domains: somatic reactions (somatic complaints, protective body postures at rest, protection of sore areas, expression, sleep pattern), psychomotor reactions (washing and/or dressing, mobility), and psychosocial reactions (communication, social life, and behavioral problems). Although this scale is recommended for use only in noncommunicative patients, several items rely on the patient’s ability to communicate. 96 The only validation study for the instrument compared it with the Visual Analog Scale (VAS), a self-report tool. 97 Thus, its utility is unclear.
The Pain Assessment in Advanced Dementia (PAINAD) scale is an observational metric based on five items: breathing, vocalization, facial expression, body language, and consolability. 98 The PAINAD scale was validated in a cohort of severely demented nursing home patients. Reliability (consistency across time and different raters) and validity (ability to measure pain and not other conditions [eg, depression] and to accurately reflect changes in pain experience) are adequate. 99
Fisher et al developed the Proxy Pain Questionnaire (PPQ) to investigate the association in nursing home patients between caregiver reports and Minimum Data Set (MDS) reports of pain. 100 The MDS is a quarterly assessment of nursing home patients that evaluates a variety of parameters, including pain. 101,102 The attractiveness of the PPQ lies in its simplicity. It consists of three questions: (1) Within the last week has this resident experienced pain? (2) How often does this resident experience pain? (3) In general, when this patient is in pain, how would you describe the extent of the pain? Despite the straightforward nature of the instrument, it is very reliable. Unfortunately, it was found to be only weakly valid in comparison with the MDS. 99 However, pain may be under-reported by use of the MDS. 103 Thus, further validation is necessary to determine the utility of the PPQ.
The Pain Assessment for the Dementing Elderly (PADE) is another instrument designed for use by non-research-trained staff. It consists of 24 brief questions and takes about 5 minutes to complete. 104 It is highly reliable, but weakly valid. 99
Modification of This Question in Light of New Research: A key aspect of perioperative care is excellent pain control. Pain is associated with a variety of adverse outcomes. 105–111 Satisfactory postoperative analgesia depends on the ability to assess level of pain. With an aging population, increasing numbers of cognitively impaired patients will require surgery. For these patients, traditional self-reported measures of pain, such as the VAS and the Numeric Pain Scale scores, clearly are inadequate. 112 In addition, existing tools are underutilized. 113 Development of instruments to assess pain in this population is essential. Although the studies described are encouraging, none was tested in a surgical population. Future development of such metrics needs to include postoperative patients in the validation cohort. In addition, when possible, instruments should be designed to be used by caregivers rather than specially trained researchers or clinicians in order to increase vastly the utility of future assessments.
Anes 29 (Level C): Improved tools for assessing pain in cognitively impaired elderly patients (Anes 28) should be used to determine the adequacy of pain management strategies in this group of patients.
New Research Addressing This Question: As noted in Anes 28, no sufficient tools have been developed.
Modification of This Question in Light of New Research: The use of new tools for evaluating pain in this population to test pain management strategies will be crucial to improving perioperative care in this vulnerable population. Anes 29 should remain on the research agenda unchanged.
Anes 30 (Level D): A retrospective review is needed to determine the incidence of polypharmacy with combinations of drugs that might contribute to complications (hypotension, bradycardia, falls, confusion, bleeding diathesis, constipation, and urinary retention) in geriatric surgical patients.
New Research Addressing This Question: No research published since the initial literature review was found.
Modification of This Question in Light of New Research: The baseline frequency of polypharmacy in the elderly person, 114,115 combined with the large number of medications administered in the perioperative period, makes answering this question important. Anes 30 should not be altered or eliminated.
Anes 31 (Level A): The effect on outcomes for elderly surgical patients of simplifying drug regimens in hospital or of communicating that information to primary care physicians should be examined in a randomized controlled trial.
New Research Addressing This Question: No research published since the initial literature review was found.
Modification of This Question in Light of New Research: Acute hospitalization provides an excellent opportunity to review and appropriately adjust the often complicated medication regimens of elderly surgical patients and thus to minimize the risks associated with polypharmacy. However, a program to do so would require a great deal of effort on the part of health care providers. Therefore, there must be studies evaluating the efficacy of this approach before it is implemented.
Anes 32 (Level B): Cross-sectional studies, with multivariate analysis, are needed to determine whether the risk factors for delirium in elderly surgical patients are the same as those for elderly medical patients.
New Research Addressing This Question: A single-center prospective cohort study addressing this issue in elective orthopedic surgical patients has recently been completed. Results are forthcoming. 116
Modification of This Question in Light of New Research: The ongoing study may clarify this issue. However, further investigations will be necessary to confirm its results and test its applicability across a broad range of elderly surgical patients.
Anes 33 (Level B): Studies are needed on the utility of the Confusion Assessment Method (CAM) for serial testing of elderly patients before and after surgery to facilitate the diagnosis of postoperative delirium. The CAM should be compared with other tests of cognitive function and with the clinical diagnosis for delirium. At the same time, since dementia is the leading predisposing factor for delirium, the utility of short mental status tests to make the preoperative diagnosis of early dementia should be tested, using a full psychiatric examination as the gold standard.
New Research Addressing This Question: The literature search did not reveal any studies assessing the utility of the CAM in a purely surgical population. This is significant for two reasons: First, the CAM was developed by the use of patients in a general medicine ward and an outpatient population. 117 Applying the instrument in a surgical population may be difficult. For example, using it in the early postoperative period may identify patients as delirious when they are still under the influence of anesthetic agents. It is unclear whether delirium in that circumstance is prognostic of outcome in the same way as delirium that occurs after the effects of anesthetics have abated. However, there is evidence that, in hip fracture surgery patients, a positive CAM 60 minutes after discontinuation of volatile anesthetic agents predicts later postoperative delirium. 118 And second, the CAM already has been used as the endpoint for a number of studies examining postoperative delirium. 13,17,23,50,90,105,118–137 Without validation and agreement on the appropriate time at which to begin screening for delirium, it is difficult to interpret the findings of studies applying the CAM in surgical patients.
The CAM has been compared with other screening tests for delirium. Gaudreau et al compared the Nursing Delirium Screening Scale (Nu-DESC) with the CAM, the Memorial Delirium Assessment Scale (MDAS), 138 and the criteria from the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV). The Nu-DESC is a brief, five-item observational instrument based on the DSM-IV criteria for delirium. In a comparison with the CAM, the Nu-DESC was found to be 85.7% sensitive and 86.8% specific. However, CAM results, not DSM-IV criteria, were considered the gold standard for the diagnosis of delirium in this study. 139 Adamis et al compared the Delirium Rating Scale, 140 a 10-item clinician-rated symptom rating scale for delirium, with the CAM and found good to very good agreement between the two instruments in acutely ill hospitalized patients, but, again, no gold standard for diagnosis was used for comparison. 141 Other studies have the same methodologic flaw. 142–144
Regarding preoperative cognitive screening, the MMSE 19 is a widely accepted screen for dementia; however, it is too cumbersome and time-consuming for use in busy clinical practices. Because of this, short tests of cognition have been developed. 145–148 The Mini-Cog is a more recently developed screening tool for dementia and is considered to be among the most useful. 149,150 It consists of three-item recall and a clock-drawing test and is to be performed in 2 to 4 minutes, rather than the 5 to 12 minutes necessary for the MMSE. It is similar in sensitivity and specificity to the MMSE and formal neuropsychologic testing for detecting dementia. 151 No studies were identified that used the Mini-Cog in a surgical population.
Modification of This Question in Light of New Research: There remains a need for studies examining the use of the CAM in elderly surgical patients. Research is needed not only to validate its usefulness in this population but also to determine the optimal time at which to apply it in relation to emergence from anesthesia. Studies comparing other delirium assessment instruments with the CAM must utilize gold standard clinical criteria, such as those in the updated text revision of DSM-IV (the DSM-IV-TR). These projects are essential to improving understanding of issues regarding postoperative delirium because without valid, standardized instruments to determine the presence of delirium, it will be difficult to interpret the results of studies.
Although research into the Mini-Cog is encouraging, there remains a need for further investigation into the utility of this and other instruments for surgical populations.
Anes 34 (Level A): Prospective controlled (nonrandomized; ie, by ward or unit) trials in patients at moderate to high risk for delirium should be performed to determine the effect of preoperative or postoperative interventions on the incidence of delirium.
New Research Addressing This Question: A longitudinal prospective sequential study examined the effect of a nurse-led interdisciplinary intervention for delirium. 136 The intervention consisted of nursing education, systematic screening for delirium using the NEECHAM Confusion Scale, 152 consultation with a trained geriatric care provider, and use of a scheduled pain medication protocol. A cohort of 60 elderly patients admitted for hip fracture treatment received standard care; 60 subsequent patients (median age 80 years, interquartile range 12 years) received the intervention. The incidence of delirium was similar in the control and intervention cohorts (23.3% versus 20.0%, respectively, P = .82). However, the duration of delirium was shorter in the intervention cohort (P = .03). Severity of delirium was reduced in the intervention group as well. In addition, cognitive function measured by the memory subdimension of the MMSE 19 was better in the delirious patients than in the interventional group. Unfortunately, the intervention did not result in improved rehabilitation or functional status.
In a prospective trial Marcantonio et al randomized 126 hip fracture patients to either usual care or to receive a proactive geriatrics consultation. 50 In the intervention group, a geriatrician made daily rounds and made recommendations based on a structured protocol (see Table 2.1). The number of recommendations was limited to five on the initial visit and three in subsequent visits in order to improve compliance. Mental status and incidence and severity of delirium were assessed daily by blinded observers using validated scales. 19,117,138,153,154 The incidence of delirium was reduced in the intervention group (32% versus 50%, P = .04, relative risk 0.64, 95% CI: 0.37 to 0.98). Delirium severity was reduced as well. This study is encouraging because the recommendations used for the intervention group could, by and large, be implemented through education of surgical care providers and adapted into care protocols for geriatric surgical patients without significant increase of health care resource utilization.
Table 2.1—Content of the Structured Geriatrics Consultation to Prevent Postoperative Delirium
1. Adequate CNS oxygen delivery: 2. Fluid/electrolyte balance: 3. Treatment of severe pain: 4. Elimination of unnecessary medicine: 5. Regulation of bowel/bladder function: 6. Adequate nutritional intake: 7. Early mobilization and rehabilitation: 8. Prevention, early detection, and treatment of major postoperative complications: 9. Appropriate environmental stimuli: 10. Treatment of agitated delirium: In addition to studies examining altered paradigms of perioperative care for geriatric surgical patients, pharmacologic prevention of delirium has been examined. In a study examining the use of low-dose haloperidol for the prevention of delirium, 430 hip surgery patients ≥ 70 years old were randomized to receive 1.5 mg per day of haloperidol or placebo starting preoperatively and continuing for up to 3 days after surgery. 155 Though there was no difference in the incidence of delirium, severity and duration were reduced in those receiving haloperidol. Importantly, hospital length of stay also was reduced (mean difference = 5.5 days, 95% CI: 1.4 to 2.3 days, P < .001). Again, this study demonstrates that a relatively simple and inexpensive intervention can markedly influence perioperative outcome.
Modification of This Question in Light of New Research: The studies described show that pre- and postoperative interventions can improve delirium-related outcomes in elderly hip fracture patients. However, they were performed only in hip surgery patients, a group at relatively high risk for postoperative delirium. Further work is required to determine whether these interventions are effective in other high- and moderate-risk surgical populations. In addition, studies are needed that address the relative contribution of each component of multimodal delirium prevention programs, such as those in the studies by Marcantonio and Milisen. 135,136 Subsequent studies can examine the efficacy of optimized multimodal protocols that are based on projects identifying the most effective individual interventions for preventing delirium. Finally, it is crucial to assess whether the recommendations provided by the geriatric consultation service in the Marcantonio study can be successfully incorporated into the routine care of geriatric surgical patients without the necessity of separate geriatric consultation. This goal is especially important, given the increasing numbers of older patients undergoing surgery and the relatively limited number of geriatricians and hospitalists available to provide such services.
New Research Addressing This Question: No research published since the initial literature review was found.
Modification of This Question in Light of New Research: Anes 35 should remain on the research agenda unchanged. Although efforts have been made to demonstrate improvements in functional status in elderly chronic pain patients that are related to opiate therapy (see Anes 36), there remains a need for hypothesis-generating studies that examine other therapies, including implantable spinal cord stimulators, pain blocks, physical and psychological rehabilitation therapy, and complementary therapies such as acupuncture.
New Research Addressing This Question: Although no hypothesis-generating studies as suggested in Anes 35 have been published, there have been a number of reports examining the effects of pain intervention on functional outcome. Specifically, these studies evaluated the impact of opiate therapy. Eight studies 156–163 published since 2001 that enrolled a total of 2640 patients were included in a recent systematic review that included three other studies, for a total of 2877 patients. The quality of the included studies ranged from low to moderate or high, but overall, opiates were associated with significant improvements in functional status, including quality of life. 164
No studies examining functional status outcomes as they relate to other pain management strategies in elderly patients were identified.
Modification of This Question in Light of New Research: Further studies are needed and should be based on outcomes from studies to address Anes 35. Here, too, a variety of therapeutic modalities should be examined.
New Research Addressing This Question: Postherpetic neuralgia (PHN) is common in the elderly age group. 165,166 The pain associated with it is associated with significant functional disability. 167 Thus, studies aimed at the prevention of PHN have considerable public health significance.
The literature search did not identify any trials published since New Frontiers comparing the use of antiviral therapy versus standard care for the prevention of PHN. A systematic review concluded that oral acyclovir provided a modest reduction in pain 1 to 3 months after zoster onset. 168
No studies exploring the use of steroids and analgesics were identified.
For theoretical reasons, combination therapy may be more effective in preventing PHN. Pasqualucci et al randomized 600 patients with acute herpes zoster to receive either intravenous acyclovir for 9 days combined with prednisolone for 21 days or epidural bupivacaine (boluses given every 6 to 12 hours) combined with epidural prednisolone every 3 to 4 days. Treatment duration in the epidural group ranged from 7 to 21 days. The incidence of PHN pain at 1 year was 22.2% and 1.6% in the acyclovir and epidural groups, respectively (P < .001). 169
Because this approach requires hospitalization and extensive use of health care resources, van Wijck et al compared oral antivirals and analgesics with a single epidural injection of methylprednisolone and bupivacaine. Although the incidence of pain at 1 month was modestly reduced in the epidural group (48% versus 58%, relative risk 0.38, 95% CI: 0.71 to 0.97, P = .02), there were no differences at 3 and 6 months. 170
A more recent approach to the prevention of PHN is the use of vaccines. With age, there is a progressive decline in cell-mediated immunity to varicella zoster virus; the incidence and severity of herpes zoster and PHN are associated with this decline. 171–179 That recurrence of herpes zoster is rare in immunocompetent individuals, presumably as a result of an immunizing effect, suggests that strategies to increase cell-mediated immunity would be useful in the prevention of PHN. 171–173,180,181
In a multicenter randomized, double-blind, and placebo-controlled trial of over 38,000 patients aged 60 years and older, Oxman et al tested an investigational live attenuated varicella zoster vaccine for the prevention of PHN. The results were promising. In patients who received the vaccine, the herpes zoster and PHN were reduced by 51.3% and 66.5%, respectively (for both P < .001). There was a low incidence of side effects. 182
Modification of This Question in Light of New Research: Given the lack of definitive answers regarding other strategies for its prevention, there remains a need for further prospective studies examining the role of antiviral, analgesic, and anti-inflammatory therapies during acute zoster in the prevention of PHN. Special consideration should be given to combination therapies and examining resource utilization. Moreover, Anes 37 should be designated as an A-level agenda item because PHN is both common and associated with significant functional disability among older persons.
As promising as vaccines may be for the prevention of PHN, in their current form they are not 100% effective in preventing PHN. Clearly, more research regarding this therapy is needed.
New Research Addressing This Question: No research published since the initial literature review was found.
Modification of This Question in Light of New Research: Elderly persons have diminished physiologic functional reserve that affects all organ systems. (See New Frontiers, pp. 17–18.) This makes them more prone to adverse side effects from chronic pain treatment modalities. Information provided by Anes 38 will allow for more rational approach to treatment decision making. This is analogous to the possibility of using functional status to assist in surgical planning (see Anes 6). For example, patients with a high propensity for confusion may not be appropriate candidates for chronic opiate therapy.
New Research Addressing This Question: No research published since the initial literature review was found.
Modification of This Question in Light of New Research: Anes 39 should remain unaltered. Given the diminished functional reserve capacity in cognitively impaired patients and potential for pain therapies (eg, opiates) to alter cognitive function, this question remains vital.
New Research Addressing This Question: No research published since the initial literature review was found.
Modification of This Question in Light of New Research: Assessment of chronic pain includes evaluating the effects of pain, including functional status and altered mood. However, in chronic pain clinical practice and research, the same tools used to assess acute pain are used for chronic pain. Therefore, information derived in Anes 28 could be applied to Anes 40.
New Research Addressing This Question: A study by Wong et al is an example of the usefulness of the research proposed in Anes 41. In this prospective, randomized, double-blind trial, 100 patients with pancreatic cancer were assigned to receive either neurolytic celiac plexus block with systemic analgesic therapy as needed or optimized systemic analgesic therapy along with a sham block. Patients were followed for 1 year or until death. Pain control (P ≤ .01) and quality of life (P < .001) improved 1 week after randomization in both groups, but there was a larger decrease in pain in those who received the block (P = .005). Pain control was better in the block group over time, as well (P = .01). However, opiate consumption, opiate side effects, and quality of life were similar between groups. There was no difference in survival between groups. 183
Modification of This Question in Light of New Research: More studies, such as the one described but focusing on elderly patients, are needed across the spectrum of chronic pain conditions and therapies. This is especially crucial because many current invasive and implantable pain therapies are costly. If they cannot be shown to confer significant clinical and functional benefit, there will be little justification for their use in treating elderly patients.
Since publication of New Frontiers, two areas of study have emerged—one new, one old—that bear further research in geriatric patients.
Diabetes mellitus is associated with adverse perioperative outcomes. 184–190 Even in patients without diabetes, hyperglycemia is associated with increased mortality, incidence of postoperative infection, and resource utilization in surgical patients. 191,192 Because of this, a number of recent studies have examined the role of glucose control in critically ill and surgical patients. 191,193–202 In 2001, van den Berghe et al published a landmark study examining the effect of intensive insulin therapy on mortality, morbidity, and resource utilization in critically ill patients. In this prospective study, patients admitted to a surgical intensive care unit were randomized to receive either intensive insulin therapy (maintenance of blood glucose between 80 and 110 mg/dL) or conventional treatment (insulin only for blood glucose > 215 mg/dL and subsequent maintenance of blood glucose between 180 and 200 mg/dL). Overall, intensive care unit mortality was 4.6% and 8.0% (P > .04) in the intensive insulin and conventional treatment groups, respectively. The adjusted reduction in mortality was 32% (95% CI: 2% to 55%). 202 Intensive insulin therapy is associated with better immune system function, 203 reduced inflammation, 204 improved lipid profiles, 205 preservation of endothelial function, 206 and protected mitochondrial structure and function. 207 However, it appears that the clinical benefit of this therapy is due to blood glucose levels independently of insulin dose. 208,209
Therefore, studies addressing the role of intra- and postoperative glucose management in elderly surgical patients are imperative. However, hypoglycemia is associated with mortality in elderly patients, 210 and the incidence of hypoglycemia is high with intensive insulin therapy. 202 Given that glucose would not likely be as well controlled outside of a study situation, 211 studies examining morbidity and mortality associated with various glucose levels should also address morbidity caused by strategies to lower perioperative glucose.
SOURCE: Marcantonio ER, Flacker JM, Wright RJ, et al. Reducing delirium after hip fracture: a randomized trial. J Am Geriatr Soc 2001;49:516–522. Adapted and reprinted with permission.
Progress in Chronic Pain
Anes 35 (Level B): The first priority in chronic pain trials is large cross-sectional studies that are powered to identify any relationship between pain intervention and functional outcomes.
Anes 36 (Level A): With the establishment of any relationship between intervention for chronic pain and functional outcomes in elderly patients (Anes 35), it must be determined prospectively if specific chronic pain therapies can improve functional outcomes in treatment groups relative to a historical or concurrent nonrandomized control.
Anes 37 (Level C): Given the high incidence rate of herpes zoster in the geriatric population, further prospective studies are needed to determine if antiviral, analgesic, or anti-inflammatory therapies during acute zoster can reduce, relative to standard care, the development of chronic postherpetic neuralgia.
Anes 38 (Level D): Cross-sectional studies documenting the association of chronic pain therapy with the incidence of complications like confusion, postural hypotension, falls, urinary retention, and constipation in the elderly population are needed.
Anes 39 (Level D): Cross-sectional studies that describe pain management in cognitively impaired patients, relative to a nonimpaired population, are needed.
Anes 40 (Level D): Pain assessment tools for chronic pain in the cognitively impaired elderly patient must be compared prospectively with standard assessment methods.
Anes 41 (Level C): Prospective trials comparing different analgesic strategies with regard to clinical and functional outcomes are needed.
New Horizons in Geriatric Anesthesia
Perioperative Glucose Control
Anes 42 (Level A): Large prospective randomized studies addressing the role of intraoperative and postoperative glucose management in elderly surgical patients are needed, to include assessment of associations of specific glucose levels or specific glucose control strategies with morbidity and mortality.
New Frontiers concluded that research on the impact of intraoperative anesthetic management on perioperative outcomes in the elderly was unlikely to be fruitful. (See New Frontiers, pp. 23–25.) Well-designed investigation into the effects on the perioperative outcomes of regional versus general anesthesia—arguably the most substantively dissimilar choices regarding intraoperative anesthetic management—failed to demonstrate clinically significant differences in outcomes between the two techniques. 212 Given the equivalency in outcome between these two vastly different techniques, it was felt that further research into this area was unlikely to yield clinically useful information and should not be pursued, especially given limited funds for research. However, recent studies suggest that anesthetic management may indeed affect cognitive outcome.
For example, in vivo studies demonstrate that exposing neurons to volatile anesthetic gases increases oligomerization and toxicity of Alzheimer-like proteins. 213 Further, in an animal model, exposure to clinically relevant concentrations of volatile anesthetics leads to impaired memory and learning. 214–216
Design issues may be why previous studies of regional versus general anesthesia failed to show differences in cognitive outcomes. In theory, the advantage of regional anesthetic techniques is that they minimize exposure to medications that act on the central nervous system and block the neuroendocrine stress response to surgery. 217 A major weakness of previous studies examining the effect of regional versus general anesthesia on cognitive outcomes was that postoperative analgesic management was not controlled. 135,218,219 Thus, patients were exposed to pharmacologic agents that may have affected cognitive outcomes (eg, opiates). It is possible that improvements in cognitive outcome related to anesthetic or analgesic technique require that those techniques be continued into the postoperative period. Recently, postoperative analgesic techniques have been developed which employ indwelling perineural catheters to deliver continuous infusions of local anesthetics. These techniques are associated with significant reductions in parenteral opiate requirements and excellent analgesia. 220 Preliminary evidence suggests that, by themselves, these techniques are associated with reduced incidence of postoperative delirium. In addition, these techniques may lead to improved cognitive outcomes when patients also receive regional instead of general anesthesia. 90
In light of the development of peripheral nerve catheter techniques and these data, prospective randomized trials should be performed that examine the effect of intravenous patient-controlled opiate analgesia versus central neuraxial or peripheral nerve catheter analgesia, or both, on cognitive outcomes in the setting of both regional versus general anesthesia. These studies will be difficult and expensive to perform, but they are exceedingly important, as the results may have a profound impact on the perioperative care of elderly surgical patients.
Anes 43 (Level A): Prospective randomized trials are needed to determine the effect of intravenous patient-controlled opiate analgesia versus central or peripheral nerve catheter analgesia, or both, on cognitive outcomes in the setting of both regional versus general anesthesia.