Correlation of postoperative epidural analgesia on morbidity and mortality after colectomy in Medicare patients

Journal of Clinical Anesthesia (2006) 18, 594 – 599

Original contribution

Correlation of postoperative epidural analgesia on morbidity and mortality after colectomy in Medicare patientsB Christopher L. Wu MD (Associate Professor)a,*, Andrew J. Rowlingson BA (Research Associate)a, Robert Herbert (Research Associate)b, Jeffrey M. Richman MD (Assistant Professor)a, Robert A.F. Andrews MD (Resident)a, Lee A. Fleisher MD (Professor)c a

Department of Anesthesiology and Critical Care Medicine, The Johns Hopkins University, Baltimore, MD 21287, USA Department of Health Policy and Management, The Johns Hopkins Bloomberg School of Public Health, The Johns Hopkins University, Baltimore, MD 21287, USA c Department of Anesthesia, University of Pennsylvania, Philadelphia, PA 19104, USA b

Received 20 September 2005; revised 24 March 2006; accepted 26 March 2006

Keywords: Epidural; Medicare; Mortality; Colectomy; Pain; Postoperative

Abstract Study Objective: To examine, with a large database, the effect of postoperative epidural analgesia (vs systemic analgesia) on mortality after colectomy is unclear. Design: Retrospective cohort (database) design. Setting: Medicare beneficiaries undergoing elective colectomy. Patients: We examined a cohort of 12 817 patients obtained from a 5% nationally random sample of Medicare beneficiaries from 1997 to 2001 who underwent elective partial excision of the large intestine. Interventions: Patients were divided into two groups depending on the presence or absence of billing for postoperative epidural analgesia (Current Procedural Terminology code 01996). Measurements: The primary outcomes assessed were death at 7 and 30 days after the procedure. The rates of major morbidity were also compared. Multivariate regression analysis incorporating race, gender, age, comorbidities, hospital size, hospital teaching status, and hospital technology status was performed to determine whether the presence of postoperative epidural analgesia had an independent effect on mortality or major morbidity. Main Results: Multivariate regression analysis revealed that there was no difference between the groups with regard to overall major morbidity; however, the presence of epidural analgesia was associated with a significantly lower odds of death at 7 days (odds ratio, 0.35; 95% confidence interval, 0.21-0.59; P b 0.0001) and 30 days (odds ratio, 0.54; 95% confidence interval, 0.42-0.70; P b 0.0001) after surgery.

Presented in part at the American Society of Regional Anesthesia 29th Annual Spring Meeting, Orlando, FL, March 2004. B Grant support: Funded by the American Society of Regional Anesthesia and Pain Medicine/Carl Koller Memorial Research Fund (CLW). * Corresponding author. Tel.: +1 410 614 0401; fax: +1 410 614 1796. E-mail address: [email protected] (C.L. Wu). 0952-8180/$ – see front matter D 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.jclinane.2006.03.020

Correlation of postoperative epidural analgesia

595

Conclusions: The presence of postoperative epidural analgesia may decrease the odds of death after elective colectomy; however, the mechanism of such a benefit is not clear from our analysis. D 2006 Elsevier Inc. All rights reserved.

1. Introduction Older patients undergoing colectomy may be at higher risk for perioperative morbidity and mortality [1]. Although many factors may influence morbidity and mortality in these patients, the perioperative use of epidural anesthesia and analgesia, through the provision of postoperative analgesia and attenuation of perioperative pathophysiology, may facilitate return of gastrointestinal function after colectomy [2-5]. However, the benefits of epidural anesthesia and analgesia on perioperative mortality after colectomy are uncertain as available randomized controlled trials (RCTs) are underpowered to detect such a relatively uncommon endpoint [6-8]. An alternate method to examine the correlation of perioperative epidural anesthesia and analgesia on perioperative mortality after colectomy is to analyze large databases, such as Medicare claims, which may contain data on millions of patients. Despite the limitations of database analysis, the acquisition and analysis of a database containing millions of patients would facilitate assessment of rare outcomes such as death and reflect actual morbidity and mortality rates from clinical practice. We previously used administrative databases to examine the effectiveness of postoperative epidural analgesia on morbidity and mortality in Medicare patients undergoing a mix of surgical procedures [9], and we currently have performed a subgroup analysis of this data to specifically examine a relatively high-risk procedure (ie, colectomy) where available RCTs indicate a potential benefit for postoperative epidural analgesia in decreasing morbidity.

2. Materials and methods A 5% nationally random sample of the Medicare population from 1997 to 2001 was obtained and used for the analysis. All patients covered by Medicare (aged z65 years or younger patients on renal dialysis or who are disabled) living in any of the 50 states or District of Columbia, and having both parts A (hospital) and B (physician) coverage, were eligible to be included in the dataset. From this population, the Medicare program creates a research database containing 5% of all Medicare beneficiaries by randomly selecting individuals based on the final two digits of the beneficiary’s social security number, with the final 2 digits of the social security number assigned randomly. This database contains a nationally representative random sample of Medicare beneficiaries and information on all inpatient hospital, skilled nursing facility,

home health care, outpatient hospital, and physician encounters incurred. Patients older than 65 years in the Medicare database who underwent colectomy were identified from part B data using the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) codes for partial excision of large intestine (45.73, 45.76). Patients were eligible for inclusion if the procedure was performed by a surgeon (part B), with an accompanying inpatient record for the same procedure (part A). Emergency surgical cases were excluded. Patients who met these criteria were divided into two groups depending on the presence of the Current Procedural Terminology code (part B) indicating presence of postoperative epidural analgesia using the Current Procedural Terminology code 01996 (daily management of epidural analgesia): the absence of a charge was assumed to indicate the absence of postoperative epidural analgesia. In an attempt to limit our analysis to epidural catheters used for postoperative analgesia, only epidurals that included a bill for management of postoperative analgesia within one day of the surgical procedure were eligible for consideration. The methodology of our current analysis was similar to that of our prior analyses [9,10]. Major morbidity at 7 and 30 days after surgery was obtained from part B for each group based on the presence of the following ICD-9-CM diagnosis codes: acute myocardial infarction (MI), 410.x; deep venous thrombosis (DVT), 453.8; pulmonary embolism (PE), 415.1; angina, 413.x; respiratory failure, 518.8; heart failure, 428.x; cardiac dysrhythmias, 427.x; pneumonia, 480.x-486.x; pulmonary edema, 518.5; sepsis, 038.x; acute renal failure (ARF), 584.x or 586.x; paralytic ileus, 560.1; transient organic syndrome, 780.02; somnolence, 789.09; and acute cerebrovascular event (CNS), 436.x; with x representing any number from 0 to 9. The number of deaths within 7 and 30 days after the procedure was also obtained. Counts for adverse events and deaths were cumulative such that complications that occurred within 7 days were also included in those at 30 days. A Deyo modification of the Charlson index, a comorbidity measure based on diagnosis subgroups developed by Elixhauser, was calculated for both groups [11-13]. These comorbidity measures have been used with large administrative datasets and have been suggested as a comorbidity measure for database analysis in studies of perioperative morbidity and mortality [14]. A multivariate model was created to control for factors [15-19] known to influence prognosis. Multivariate regression analysis incorporating race, gender, age, comorbidities (for both modified Charlson and Elixhauser indices separately), hospital size, hospital teaching status, and hospital

596

C.L. Wu et al.

Table 1 Demographic data (Medicare 1997-2001): partial excision of large intestine (excluding emergencies)

Gender (M/F) Race White African-American Other Age 65-69 y 70-74 y 75-79 y 80-84 y N85 y Charlson Index

No epidural (n = 10 564)

Epidural (n = 2253)

P

4208/6356 (40%/60%)

963/1290 (43%/ 57%)

0.67

9642 (91%) 639 (6%) 283 (3%)

2116 (93%) 101 (4%) 36 (2%)

1986 2723 2546 1985 1324

(19%) (26%) (24%) (19%) (13%)

0.2 F 0.7

467 557 582 422 225

Outcome

(21%) (25%) (26%) (19%) (10%)

0.2 F 0.6

0.74

0.96

1.0

technology status (presence of transplant and coronary artery bypass services) was performed to determine whether the presence of postoperative epidural analgesia had an independent effect on mortality or major morbidity. Multivariate logistic models were developed separately for death at 7 days, death at 30 days, and each of the above-listed complications at 7 and 30 days. Logistic regression equations were calculated using SAS version 8.02 (SAS Institute, Cary, NC). v 2 Test was used to analyze difference in the proportions of age, race, and gender, and t test was used to compare lengths of stay and illness severity Table 2 Unadjusted 7-day morbidity and mortality data (Medicare 1997-2001): partial excision of large intestine (excluding emergencies)

Death Acute MI Angina Dysrhythmias Heart failure Pneumonia Pulmonary edema Respiratory failure DVT PE Sepsis ARF CNS Transient CNS Paralytic ileus Any adverse P b 0.05.

No epidural (n = 10 564)

Epidural (n = 2253)

Totals

Rate/1000

Totals

Rate/1000

283 1 0 4 0 2 4 1 0 1 2 1 0 0 0 16

26.79 0.09 0 0.38 0 0.19 0.38 0.09 0 0.09 0.19 0.09 0 0 0 1.51

21 0 0 0 0 0 0 0 0 0 2 0 0 0 0 2

9.32 0 0 0 0 0 0 0 0 0 0.89 0 0 0 0 0.89

7 days OR (CI)

Death

Charlson Index data presented as means F SD. F indicates female; M, male.

Variable

Table 3 Adjusted association between postoperative epidural analgesia and outcome at 7 and 30 days after partial excision of large intestine (excluding emergencies)

Acute MI

30 days P

OR (CI)

0.35 b0.0001 0.54 (0.21-0.59) (0.42-0.70) * 0.95 1.07 (0.23-5.0) N/M * * 0.96 *

Angina Cardiac dysrhythmias Heart failure N/M Pneumonia *

0.95

Pulmonary edema Respiratory failure DVT

*

0.94

*

0.96

PE

*

0.94

Sepsis

*

0.95

Acute renal failure CNS Transient organic syndrome Paralytic ileus

*

0.95

N/M

Any adverse

*

N/M

N/M N/M

0.95

P b0.0001 0.93 0.94 0.95

* 0.36 (0.05-2.8) *

0.948 0.32

0.79 (0.18-3.54) 1.08 (0.31-3.83) 1.83 (0.37-9.23) 2.20 (0.83-5.85) 0.93 (0.11-7.85) * *

0.76

2.12 (0.20-22.43) 0.72 (0.44-1.19)

0.53

0.96

0.90 0.46 0.11 0.95 0.96 0.93

0.21

N/M indicates not meaningful. * OR b0.001 or N999.999 with CI b0.001 and N999.999.

measures between the groups. All 7- and 30-day morbidity and mortality data are reported as an odds ratio (OR) with 95% confidence intervals (CI). A P b 0.05 was defined as a statistically significant difference.

3. Results From the 5% random sample of 1997 to 2001 Medicare claims data, we identified 2253 patients who underwent partial excision of large intestine and received postoperative epidural analgesia; there were 10 564 patients who underwent these procedures without epidural analgesia. Demographic data are shown in Table 1. There were no differences between the groups with regard to gender, race, age, or illness severity (eg, Charlson index). The unadjusted data for 7-day mortality and morbidity are shown in Table 2. The overall unadjusted 7-day mortality rate was 2.4% (2.7% for those without epidural

Correlation of postoperative epidural analgesia analgesia and 0.9% for those with epidural analgesia). Table 3 shows the adjusted data for both 7- and 30-day mortality and morbidity. Multivariate regression analysis revealed that there was no difference between the groups with regard to overall major morbidity; however, the presence of epidural analgesia was associated with a significantly lower odds of death at 7 days (OR, 0.35; 95% CI, 0.21-0.59; P b 0.0001) and 30 days (OR, 0.54; 95% CI, 0.42-0.70; P b 0.0001) after surgery. The OR of all combined morbidity and death for both 7 and 30 days after surgery increased significantly with increasing age and female gender (ie, there were no other significant correlations found). There were no differences in regression analyses when substituting the modified Charlson for the Elixhauser morbidity index. All data shown have been calculated using the Elixhauser index.

4. Discussion We performed an analysis using a sample of the Medicare claims data from 1997 to 2001 and found that the presence of epidural analgesia (vs no epidural) was associated with significantly lower odds of death at 7 and 30 days after partial excision of large intestine; however, the reasons for this decrease were not obvious from our analysis as there were no differences in the adjusted OR in the incidence of major morbidity between those who received postoperative epidural analgesia and those who did not. This is one of the largest cohorts to examine this issue, as a 5% random sample of the Medicare database from 1997 to 2001 may represent approximately 256 000 patients (45 000 with postoperative epidural analgesia and 211 000 without epidural analgesia) who underwent partial excision of large intestine during this period. Our study showed that perioperative epidural analgesia is associated with a statistically significant decrease in postoperative mortality although one of the largest available meta-analysis of randomized trials examining the efficacy of neuraxial (epidural and spinal) anesthesia and analgesia on perioperative morbidity and mortality found no difference in perioperative mortality for abdominal surgery [20]. Individual RCTs in subjects undergoing colectomy (typically b200 subjects per study) do not demonstrate a benefit of epidural anesthesia and analgesia in decreasing perioperative mortality [2-5]; however, these RCTs are underpowered for assessing a rare outcome such as death and were not designed to assess the efficacy of epidural techniques on mortality [6]. For instance, the number of subjects needed for an RCT to determine whether epidural anesthesia and analgesia would have an effect on mortality in a high-risk surgical group (eg, vascular patients) would be approximately 24 000 patients total [7], whereas an RCT in a lowerrisk group (eg, total hip replacement) would necessitate the enrollment of 1.2 million subjects [8]. The decrease in postoperative mortality associated with perioperative epidural analgesia in our study was not

597 coupled with a concomitant decrease in the perioperative morbidities, reflecting one of the limitations of using databases for research (please see below for additional details describing the limitations of our work). As such, it is unclear whether the many physiologic benefits of epidural analgesia [2,6,21-25] would have contributed to a decrease in postoperative mortality resulting from a decrease in perioperative morbidities. Available data suggest that perioperative use of epidural analgesia is associated with a decrease in pulmonary, cardiovascular, gastrointestinal, and coagulation-related complications. A systematic review of RCTs demonstrates that perioperative epidural analgesia facilitates return of gastrointestinal function after abdominal surgery [3]. Meta-analyses and RCTs also indicate that perioperative epidural analgesia may decrease the incidence of pulmonary complications [26], MI (for thoracic epidural analgesia only) [27], and coagulation-related events [6]. Unlike that seen with our prior analysis [10], there was no increase in DVT or PE associated with the presence of postoperative epidural analgesia that may be related to avoidance of systemic thromboprophylaxis (eg, low molecular weight heparins), although we would not be able to determine the presence of systemic thromboprophylaxis in our dataset. Although there were no differences between the groups with regard to the incidence of perioperative morbidity, we did note that the adjusted 7- and 30-day mortality rates were significantly lower for patients with epidural analgesia. Our mortality rates should be reasonably accurate as death is captured in the Medicare claims database where mortality is determined by looking at a date of death field in the Medicare beneficiary enrollment file which is obtained from the Social Security Administration on an ongoing basis [28,29]. The robustness of our mortality data is reflected by the fact that our mortality rates are similar to those found in other large-scale observational studies. For example, our 30-day mortality rates of 2.4% and 6.9% are similar to those reported from other large-scale studies (5.7%-6.7%) [30-33]. In addition, our 7-day mortality of 2.4% is comparable to the in-hospital mortality rates described elsewhere (2.5%-3.4%) [1,34,35]. There are several limitations of our study. The results obtained from database analysis may only suggest an association, not a causal relationship, between the variable and outcome of interest (as might be determined by an RCT). Although we attempted to account for many of the recognized factors that may affect postoperative morbidity and mortality, we did not incorporate our factors that may potentially influence outcome such as surgeon volume [36], surgeon specialty training [37], or presence of an acute pain service [38,39] in our regression model. It is possible that subsequent analyses that are able to adjust for specific patient-, hospital-, and surgeon-level variables may result in different findings than ours. In addition, epidural analgesia is not a single generic entity and how an epidural catheter is managed (ie, analgesic regimen, duration of epidural usage,

598 location of catheter placement in relationship to surgical incision, or use as part of a multimodal fashion [4,6,40]) may influence outcome; however, we were unable to extract these data from the Medicare claims. Also, we were unable to determine the presence of the use of accelerated recovery protocols, which may also influence morbidity or mortality. This reflects one of the limitations of database analysis; that is, many available databases are created mainly for administrative and billing, not research, purposes. Finally, in contrast to prospective trials where definitions for complications can be predefined, complication rates may not be accurate in database analysis as abstraction of complications from databases, most of which allow only 5 diagnoses to be entered [41], is retrospective and may be beset by ambiguities in coding, misclassification, and uncoded factors such as disease severity [42]. Complications and comorbid conditions may be underreported in databases [43,44]. For instance, approximately 20% to 30% of surgical and medical patients lack objective confirmatory evidence in the medical record to support the coding of certain complications such as postoperative pneumonia [45]. We were unable to audit the original patient records to confirm the presence of complications or specific characteristics of epidural analgesia. The presence of some of these problems can be seen in our data and caution may be warranted in interpreting the morbidity data and the effect of epidural analgesia on incidence of postoperative morbidities. For instance, the rate of both individual and aggregate morbidities at both 7 and 30 days after surgery (Tables 2 and 3) is lower than the rate of mortality, a finding that is inconsistent with that found in the literature (ie, the incidence of morbidity is generally higher than that for mortality) [30,46,47], and the incidences of major morbidities in our trial are generally lower than would be expected. Analysis of data after 2001 may provide different results (eg, lower mortality) than what we have obtained. Despite these limitations, there are some advantages for using databases (compared with RCTs) in examining this issue. As previously mentioned, an appropriately sized RCT would require an extremely large sample size, which would make such an RCT impractical, costly, and time-consuming to undertake. Database analysis is an alternate method of examining the same issue but in a more cost- and timeefficient manner, with a greater sample size and stronger external validity (generalizability). In summary, we analyzed Medicare patients undergoing partial excision of large intestine, stratified according to the presence of postoperative epidural analgesia, and found that the presence of epidural analgesia was associated with a significant decrease in 7- and 30-day mortality, although there was not a corresponding decrease in major morbidity. The results we obtained from our observational trial reflect actual clinical practice, although there are some methodological issues present. At the individual level, the results from our study may provide additional data on the general

C.L. Wu et al. benefits of postoperative epidural analgesia; however, the use of perioperative analgesia should be assessed on an individual basis [48,49], weighing the risks [50-52] and other benefits [53,54] of this technique.

Acknowledgments The authors would like to thank Professor Gerard F Anderson (Department of Health Policy and Management, The Johns Hopkins Bloomberg School of Public Health, The Johns Hopkins University, Baltimore, MD) for his guidance in our research.

References [1] Dimick JB, Cowan JA Jr, Upchurch GR Jr, Colletti LM. Hospital volume and surgical outcomes for elderly patients with colorectal cancer in the United States. J Surg Res 2003;114:50 - 6. [2] Liu S, Carpenter RL, Mackey DC, et al. Effects of perioperative analgesic technique on rate of recovery after colon surgery. Anesthesiology 1995;83:757 - 65. [3] Jorgensen H, Wetterslev J, Moiniche S, Dahl JB. Epidural local anaesthetics versus opioid-based analgesic regimens on postoperative gastrointestinal paralysis, PONV and pain after abdominal surgery. Cochrane Database Syst Rev(4):CD001893. [4] Hodgson PS, Liu SS. Thoracic epidural anaesthesia and analgesia for abdominal surgery: effects on gastrointestinal function and perfusion. Balliere’s Clin Anesth 1999;13:9 - 22. [5] Mann C, Pouzeratte Y, Boccara G, et al. Comparison of intravenous or epidural patient-controlled analgesia in the elderly after major abdominal surgery. Anesthesiology 2000;92:433 - 41. [6] Wu CL, Fleisher LA. Outcomes research in regional anesthesia and analgesia. Anesth Analg 2000;91:1232 - 42. [7] Bode RH Jr, Lewis KP, Zarich SW, et al. Cardiac outcome after peripheral vascular surgery. Comparison of general and regional anesthesia. Anesthesiology 1996;84:3 - 13. [8] Liu SS. Bank robbers and outcomes research. Reg Anesth Pain Med 2003;28:262 - 4. [9] Wu CL, Hurley RW, Anderson GF, Herbert R, Rowlingson AJ, Fleisher LA. Effect of postoperative epidural analgesia on morbidity and mortality following surgery in Medicare patients. Reg Anesth Pain Med 2004;29:525 - 33. [10] Wu CL, Anderson GF, Herbert R, Lietman SA, Fleisher LA. Effect of postoperative epidural analgesia on morbidity and mortality after total hip replacement surgery in Medicare patients. Reg Anesth Pain Med 2003;28:271 - 8. [11] Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373 - 83. [12] Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 1992;45:613 - 9. [13] Elixhauser A, Steiner C, Harris DR, Coffet RM. Comorbidity measures for use with administrative data. Med Care 1998;36:8 - 21. [14] Silber JH. Medicare claims and anesthesia clinical research: a perfect match. Reg Anesth Pain Med 2003;28:259 - 61. [15] Schrag D, Cramer LD, Bach PB, Cohen AM, Warren JL, Begg CB. Influence of hospital procedure volume on outcomes following surgery for colon cancer. JAMA 2000;284:3028 - 35. [16] Norman PE, Semmens JB, Lawrence-Brown M, Holman CD. The influence of gender on outcome following peripheral vascular surgery: a review. Cardiovasc Surg 2000;8:1111 - 5.

Correlation of postoperative epidural analgesia [17] Kennedy BS. Does race predict short-term mortality after carotid surgery? The results of a meta-analysis. J Natl Med Assoc 2002; 94:25 - 30. [18] Taylor DH, Whellan DJ, Sloan FA. Effects of admission to a teaching hospital on the cost and quality of care for Medicare beneficiaries. N Engl J Med 1999;340:293 - 9. [19] Birkmeyer JD, Siewers AE, Finlayson EV, et al. Hospital volume and surgical mortality in the United States. N Engl J Med 2002; 346:1128 - 37. [20] Rodgers A, Walker N, Schug S, et al. Reduction of postoperative mortality and morbidity with epidural or spinal anaesthesia: results from overview of randomised trials. BMJ 2000;321:1493 - 6. [21] Neal JM, Wilcox RT, Allen HW, Low DE. Near-total esophagectomy: the influence of standardized multimodal management and intraoperative fluid restriction. Reg Anesth Pain Med 2003;28:328 - 34. [22] Foss NB, Kristensen MT, Kristensen BB, Jensen PS, Kehlet H. Effect of postoperative epidural analgesia on rehabilitation and pain after hip fracture surgery: a randomized, double-blind, placebo-controlled trial. Anesthesiology 2005;102:1197 - 204. [23] Reuben SS. Preventing the development of complex regional pain syndrome after surgery. Anesthesiology 2004;101:1215 - 24. [24] Liu SS. Anesthesia and analgesia for colon surgery. Reg Anesth Pain Med 2004;29:52 - 7. [25] Katz J, Cohen L. Preventive analgesia is associated with reduced pain disability 3 weeks but not 6 months after major gynecologic surgery by laparotomy. Anesthesiology 2004;101:169 - 74. [26] Ballantyne JC, Carr DB, deFerranti S, et al. The comparative effects of postoperative analgesic therapies on pulmonary outcome: cumulative meta-analyses of randomized, controlled trials. Anesth Analg 1998; 86:598 - 612. [27] Beattie WS, Badner NH, Choi P. Epidural analgesia reduces postoperative myocardial infarction: a meta-analysis. Anesth Analg 2001;93:853 - 8. [28] Wentworth DN, Neaton JD, Rasmussen WL. An evaluation of the Social Security Administration master beneficiary record file and the National Death Index in the ascertainment of vital status. Am J Public Health 1983;73:1270 - 4. [29] Hill ME, Rosenwaike I. The Social Security Administration’s Death Master File: the completeness of death reporting at older ages. Soc Secur Bull 2002;2001-64:45 - 51. [30] Longo WE, Virgo KS, Johnson FE, et al. Risk factors for morbidity and mortality after colectomy for colon cancer. Dis Colon Rectum 2000;43:83 - 91. [31] Goodney PP, Siewers AE, Stukel TA, Lucas FL, Wennberg DE, Birkmeyer JD. Is surgery getting safer? National trends in operative mortality. J Am Coll Surg 2002;195:219 - 27. [32] Marusch F, Koch A, Schmidt U, et al. Impact of age on the short-term postoperative outcome of patients undergoing surgery for colorectal carcinoma. Int J Colorectal Dis 2002;17:177 - 84. [33] Ansari MZ, Collopy BT, Hart WG, Carson NJ, Chandraraj EJ. Inhospital mortality and associated complications after bowel surgery in Victorian public hospitals. Aust N Z J Surg 2000;70:6 - 10. [34] Alves A, Panis Y, Mathieu P, Mantion G, Kwiatkowski F, Slim K, Association Francaise de Chirurgie. Postoperative mortality and morbidity in French patients undergoing colorectal surgery: results of a prospective multicenter study. Arch Surg 2005;140:278 - 83.

599 [35] Guller U, Jain N, Hervey S, Purves H, Pietrobon R. Laparoscopic vs open colectomy: outcomes comparison based on large nationwide databases. Arch Surg 2003;138:1179 - 86. [36] Birkmeyer JD, Stukel TA, Siewers AE, Goodney PP, Wennberg DE, Lucas FL. Surgeon volume and operative mortality in the United States. N Engl J Med 2003;349:2117 - 27. [37] Callahan MA, Christos PJ, Gold HT, Mushlin AI, Daly JM. Influence of surgical subspecialty training on in-hospital mortality for gastrectomy and colectomy patients. Ann Surg 2003;238:629 - 36. [38] Carr DB, Reines HD, Schaffer J, Polomano RC, Lande S. The impact of technology on the analgesic gap and quality of acute pain management. Reg Anesth Pain Med 2005;30:286 - 91. [39] Rosenquist RW, Rosenberg J. Postoperative pain guidelines. Reg Anesth Pain Med 2003;28:279 - 88. [40] Carli F, Kehlet H. Continuous epidural analgesia for colonic surgery—but what about the future? Reg Anesth Pain Med 2005; 30:140 - 2. [41] Iezzoni LI, Foley SM, Daley J, Hughes J, Fisher ES, Heeren T. Comorbidities, complications, and coding bias. Does the number of diagnosis codes matter in predicting in-hospital mortality? JAMA 1992;267:2197 - 203. [42] Myers ER, Steege JF. Risk adjustment for complications of hysterectomy: limitations of routinely collected administrative data. Am J Obstet Gynecol 1999;181:567 - 75. [43] Romano PS, Chan BK, Schembri ME, Rainwater JA. Can administrative data be used to compare postoperative complication rates across hospitals? Med Care 2002;40:856 - 67. [44] Quan H, Parsons GA, Ghali WA. Validity of information on comorbidity derived from ICD-9-CM administrative data. Med Care 2002;40:675 - 85. [45] McCarthy EP, Iezzoni LI, Davis RB, et al. Does clinical evidence support ICD-9-CM diagnosis coding of complications? Med Care 2000;38:868 - 76. [46] Dimick JB, Pronovost PJ, Cowan Jr JA, Lipsett PA. Postoperative complication rates after hepatic resection in Maryland hospitals. Arch Surg 2003;138:41 - 6. [47] Bufalari A, Ferri M, Cao P, Cirocchi R, Bisacci R, Moggi L. Surgical care in octogenarians. Br J Surg 1996;83:1783 - 7. [48] Bergqvist D, Wu CL, Neal JM. Anticoagulation and neuraxial regional anesthesia: perspectives. Reg Anesth Pain Med 2003; 28:163 - 5. [49] Yeager MP, Carli F. Anesthesia and surgical outcomes: an orphean ambition. Reg Anesth Pain Med 2004;29:515 - 9. [50] Zink W, Graf BM. Local anesthetic myotoxicity. Reg Anesth Pain Med 2004;29:333 - 40. [51] Wong CA. Neurologic deficits and labor analgesia. Reg Anesth Pain Med 2004;29:341 - 51. [52] Moen V, Dahlgren N, Irestedt L. Severe neurological complications after central neuraxial blockades in Sweden 1990-1999. Anesthesiology 2004;101:950 - 9. [53] Wu CL, Hsu W, Richman JM, Raja SN. Postoperative cognitive function as an outcome of regional anesthesia and analgesia. Reg Anesth Pain Med 2004;29:257 - 68. [54] Block BM, Liu SS, Rowlingson AJ, Cowan AR, Cowen JA, Wu CL. Efficacy of postoperative epidural analgesia versus systemic opioids: a meta-analysis. JAMA 2003;290:2455 - 63.