- Email: [email protected]
Factors Affecting Length of Stay Following Colonic Resection
Journal of Surgical Research 146, 195–201 (2008) doi:10.1016/j.jss.2007.08.015
Factors Affecting Length of Stay Following Colonic Resection Thomas M. Schmelzer, M.D., Gamal Mostafa, M.D.,1 Amy E. Lincourt, Ph.D., Steven M. Camp, M.D., Kent W. Kercher, M.D., Timothy S. Kuwada, M.D., and B. Todd Heniford, M.D. Department of Surgery, Carolinas Medical Center, Charlotte, North Carolina Submitted for publication February 13, 2007
Introduction. In-hospital length of stay (LOS) has become a valuable measure of outcomes following any operation, which also directly impacts cost. The aim of this study was to identify the factors that affect LOS after colonic resection. Materials and methods. A retrospective analysis was performed of adult patients who underwent colonic resection over an 8-y period at a tertiary institution. Data collected included demographics, American Society of Anesthesiologists (ASA) score, preoperative comorbidities and medications, operative management, postoperative morbidity and mortality, and LOS. Statistical analysis included descriptive statistics and multiple logistic regression to identify variables predictive of prolonged LOS. Results and discussion. A total of 899 consecutive patients were identified. One hundred eighty-seven resections were performed urgently, and 712 were elective. Two-hundred forty-five cases were performed laparoscopically. Complications occurred in 205 cases (23%), and there were 32 deaths (4%). The median LOS was 7 d. Logistic analysis showed 15 variables to be predictive of prolonged LOS. These included advanced age, warfarin sodium use, ASA score >3, alcoholism, chronic obstructive pulmonary disease, end-stage renal disease, illicit drug use, total colectomy (versus segmental), open resections (versus laparoscopic), and postoperative complications. In addition, the presence of at least one postoperative complication was predictive of prolonged LOS (P ⴝ 0.0002, OR 2.4 95% CI 1.5–3.8). Conclusions. ASA score and the incidence of postoperative complications are the only significant categories of variables that predict prolonged LOS after co-
1
To whom correspondence and reprint requests should be addressed at Department of Surgery/MEB 601, Carolinas Medical Center, 1000 Blythe Boulevard, Charlotte, NC 28203. E-mail: Gamal.Mostafa@ carolinashealthcare.org.
lectomy. Laparoscopic approach and the extent of the resection are predictive as well. © 2008 Elsevier Inc. All rights reserved.
Key Words: adult; colectomy; colon resection; human; laparoscopic colectomy; length of stay; morbidity; multiple linear regression; risk factors.
INTRODUCTION
Hospital length of stay (LOS) after major surgery is becoming an increasingly important measurement of surgical outcome. The influence of the insurance industry has led to a shift toward more cost-effective management of postoperative patients, which includes early release from the hospital. Surgeons are evaluated based on their ability to deliver quality care while minimizing unnecessary costs. The impact of this shift can be seen in colonic resections. This relatively common operation was frequently associated with a 6 to 12 d postoperative hospital LOS [1]. Strict postoperative protocols have been developed to “fast track” the postoperative course with the goal LOS being 2 d [2]. The methods used to decrease the recovery time include using laparoscopic techniques, early postoperative feeding, epidurals for pain management, and early mobilization [1–5]. These methods attempt to address the common limiting factors for most patients, which are pain and bowel function; however, they do not address the issues of increased LOS due to postoperative morbidity. This study examines the potential impact of various demographics, preoperative morbidities, operative details, and postoperative morbidities on hospital LOS after colonic resection. Identifying risk factors for extended LOS can facilitate modification to the preoperative preparation and postoperative management of patients undergoing colonic resections, thereby leading to more cost-effective delivery of surgical care.
195
0022-4804/08 $34.00 © 2008 Elsevier Inc. All rights reserved.
196
JOURNAL OF SURGICAL RESEARCH: VOL. 146, NO. 2, MAY 15, 2008
METHODS After obtaining Institutional Review Board approval, a retrospective review of the medical records of adult patients (age ⬎ 18 y) undergoing a colonic resection was conducted at a single tertiary institution over an 8-y period (January 1998 to December 2005). Records were identified using ICD-9 codes for colonic resection. Trauma patients were excluded from the study. Six categories of variables were used in data collection, including demographics, preoperative comorbidities, preoperative medications, American Society of Anesthesiologist (ASA) score, operative details, and postoperative morbidity/mortality. All independent variables in these categories are shown in Table 1. For this study, urgent operations were defined as operations performed within 24 h of admission or the time of surgical consultation. Comorbidity was defined as illnesses or conditions that existed before the operation and not related to or caused by the colonic pathology. A right colectomy was defined as a resection of the colon proximal to the middle colic artery; a left colectomy as a resection of the colon distal to the middle colic artery; a total colectomy as a resection of the full length of the colon from the terminal ileum to the rectum. Analgesia used referred to the amount of postoperative narcotic used by patients (measured in morphine equivalents). A wound infection was defined as erythema of the skin incision or culture-proved wound drainage requiring antibiotics or local wound care intervention. Wound dehiscence was defined as disrup-
TABLE 1 Data Evaluated in 899 Patients Undergoing Colectomy Demographics Age BMI Comorbidities Alcoholism Asthma CHF Cirrhosis/liver disease COPD Coronary artery disease Cerebrovascular accident Diabetes Preoperative medications Antiplatelet (aspirin or clopidogrel) Warfarin sodium ASA Score Operative Information Analgesia Blood transfusion Estimated blood loss Operative time Pathologic diagnosis Postoperative morbidity Anastomotic leak Atelectasis Bleeding Cardiac events (MI, failure, dysrhythmia) Cerebrovascular accident Deep vein thrombosis Ileus Intra-abdominal abscess Pneumonia
Race Sex End stage renal disease HIV Hypertension Illicit drug use Morbid obesity Previous intra-abdominal surgery Renal insufficiency (Cr. ⬎ 1.3) Smoking Steroids
Procedure (left/right/segmental/ total colectomy) Surgical technique (laparoscopic versus open) Urgency (urgent or elective) Pulmonary embolism Renal failure Respiratory failure Sepsis Small bowel obstruction Urinary tract infection Wound dehiscence Wound infection
BMI ⫽ body mass index; CHF ⫽ congestive heart failure; COPD ⫽ chronic obstructive pulmonary disease; MI ⫽ myocardial infarction.
tion of the fascial closure resulting in exposed viscera and requiring surgical intervention. An abdominal computed tomography scan was used to diagnose intra-abdominal abscess. A computed tomography scan with oral or rectal contrast was used to diagnose or exclude anastomotic leak. Clinically significant ileus was defined as requiring nasogastric tube decompression beyond postoperative day 5. Atelectasis was recorded based on radiographic reports from chest X-rays. Bleeding was defined as a drop in hemoglobin of 2 g/dL or requiring the transfusion of two or more units of packed red blood cells within the first 48 h after colectomy. Renal failure was defined as an increase in serum creatinine by a factor of 3, a decrease in glomerular filtration by 75%, or urine output less than 0.3 mL/kg/h for a period of 24 h. Respiratory failure was defined as requiring intubation and mechanical ventilation. Cardiac events were determined by cardiac enzyme values and electrocardiogram findings. Small bowel obstruction was defined by abdominal distention with or without obstipation and radiographic findings consistent with obstruction. The primary end point for this study was postoperative hospital LOS measured in days. Data were analyzed using standard statistical methods. Descriptive statistics including means and standard deviations or counts and percentages were used to describe the study population on all variables. Differences between the prolonged LOS and nonprolonged LOS groups were calculated using Wilcoxon for continuous variables and -square or Fisher’s exact for categorical variables. Significant variables were then selected using backward elimination to determine significant independent predictors of prolonged LOS using the Wald -square test logistic regression analysis. A prolonged LOS was defined as greater than the median LOS for the group, since the LOS was not normally distributed. Comparisons were also made between the median LOS for the first 25% of the observations and the most recent 25% of observations to determine differences over the time of the study. A P-value ⬍0.05 was considered significant for all determinations. SAS System software, version 8.02 (SAS Institute, Cary, NC) was used for statistical analyses.
RESULTS
During the study period, 899 colonic resections were identified. There were 449 males and 450 female patients. The average age of the patients was 61.8 y (range 19 to 92 years). One-hundred eighty-seven resections (21%) were performed urgently, and 712 (79%) were elective. Two-hundred forty-five cases (27%) were completed laparoscopically, 45 (5%) were started laparoscopically and converted to open, and 609 (68%) were open colon resections. Cancer was the most common indication for colectomy (52%), followed by diverticular disease (34%), and colitis (9%). Four-hundred thirtythree (48%) of the resections were right colectomies, 407 (45%) were left colectomies, and there were 59 (7%) total colectomies. Complications were reported in 205 cases (23%), and there were 32 deaths (4%). The median LOS was 7 d (range 1 to 138). Descriptive statistics for the prolonged versus nonprolonged LOS groups can be seen in Table 2. Twenty-four variables were statistically different between the prolonged LOS and nonprolonged LOS groups. Logistic regression analysis of these variables showed 15 variables to be predictive of prolonged LOS (Table 3). The presence of at least one postoperative complication was also a significant predictor of prolonged LOS (P ⫽ 0.0002, OR ⫽ 2.4 95% CI 1.5–3.8).
SCHMELZER ET AL.: FACTORS AFFECTING LENGTH OF STAY AFTER COLON RESECTION
197
TABLE 2 Descriptive Statistics for All Variables Based on Prolonged versus Nonprolonged Hospital Length of Stay Nonprolonged LOS (N ⫽ 375) Demographics Mean age (years) BMI (kg/m 2) Race (Caucasian/African American/Hispanic) Sex (male/female) Comorbidities (overall) Alcoholism Asthma CHF Cirrhosis/liver disease COPD Coronary artery disease CVA Diabetes End stage renal disease HIV Hypertension Illicit drug use Morbid obesity Renal insufficiency (Cr. ⬎ 1.3) Smoking Preoperative medications Antiplatelet (aspirin or clopidogrel bisulfate) Warfarin sodium Steroids ASA score (⬍3/ⱖ3) Operative information Analgesia (epidural) Blood transfusion Estimated blood loss (mL) Operative time (min) Pathologic diagnosis (malignant/benign) Procedure (left/right/total colectomy) Surgical technique (laparoscopic versus open) Urgency (urgent or elective) Postoperative morbidity (at least one) Anastomotic leak Atelectasis Bleeding Cardiac events (MI, failure, dysrhythmia) CVA Deep vein thrombosis Ileus Intra-abdominal abscess Pneumonia Pulmonary embolism Renal failure Respiratory failure Sepsis Small bowel obstruction Urinary tract infection Wound dehiscence Wound infection
Prolonged LOS (N ⫽ 524)
P-value
57 28.1 298/66/5 178/197 204 6 21 7 1 14 32 5 97 2 0 136 2 13 11 46
62 26.8 268/133/6 271/253 302 24 7 14 3 57 93 15 162 20 3 233 11 10 27 57
⬍0.0001 0.02 0.09 0.2 0.9 0.05 ⬍0.0001 0.4 0.6 ⬍0.0001 ⬍0.0001 0.2 0.1 0.001 0.3 0.01 0.1 0.1 0.2 0.5
5 4 11 266/109
10 19 13 231/293
0.1 ⬍0.0001 0.5 ⬍0.0001
14 7 156 177 167/208 175/186/14 208/167 39/336 61 1 23 4 0 0 0 10 0 1 2 0 1 3 3 2 0 14
29 39 208 177 299/225 232/247/45 95/429 147/377 144 4 34 6 12 1 12 37 6 6 3 6 14 10 15 16 3 17
0.2 ⬍0.0001 0.8 0.9 0.0002 0.02 ⬍0.0001 ⬍0.0001 ⬍0.0001 0.2 1 1 0.002 1 0.002 ⬍0.004 0.05 0.2 1 0.04 0.001 0.2 0.03 0.005 0.2 1
BMI ⫽ body mass index; CHF ⫽ congestive heart failure; COPD ⫽ chronic obstructive pulmonary disease; CVA ⫽ cerebrovascular accident; MI ⫽ myocardial infarction.
The only preoperative category of variables that was predictive of prolonged LOS was ASA score ⱖ 3 (P ⫽ 0.003, OR ⫽ 3.5; 95% CI 1.6 –9.3). Demographics, comorbidities, preoperative medications, and operative
variables were not categorically predictive of prolonged LOS. Age was a significant predictor of prolonged LOS, and patients with advanced age were increasingly more likely to have a prolonged LOS (Table 4). Warfarin
198
JOURNAL OF SURGICAL RESEARCH: VOL. 146, NO. 2, MAY 15, 2008
TABLE 3 Significant Predictors of Hospital Length of Stay Identified by Logistic Regression P-value Demographics Age Comorbidities (at least one) Alcoholism COPD ESRD Illicit drug use Preoperative medications Warfarin sodium ASA grade (ⱖ3) Operative information Procedure (total colectomy) Surgical technique (open approach) Postoperative morbidity Cardiac failure DVT Intra-abdominal abscess Renal failure Respiratory failure Small bowel obstruction
Odds ratio
95% CI
⬍0.0001 0.3* 0.01 0.004 0.03 ⬍0.0001
N/A
N/A
2.2 3.1 6.2 10.0
1.2–4.1 1.4–6.7 1.2–33.3 4.1–24.4
0.04 0.003
6.0 3.5
1.1–26.7 1.6–9.3
0.02
4.2
1.4–13.3
⬍0.0001 0.0002 0.006 ⬍0.0001 0.002 0.02 0.02 0.04
5.3 2.4 2.8 2.6 2.9 3.3 3.7 6.0
3.4–8.5 1.5–3.8 1.3–6.5 1.6–4.1 1.5–6.0 1.3–12.5 1.8–9.3 1.1–33.3
CI ⫽ confidence interval; COPD ⫽ chronic obstructive pulmonary disease; ESRD ⫽ end-stage renal disease. * Nonsignificant value.
sodium was the only preoperative medication predictive of a prolonged LOS (P ⫽ 0.04). Comorbidities serving as significant predictors of LOS were alcoholism, chronic obstructive pulmonary disease, end-stage renal disease, and illicit drug abuse. The operative variables total colectomy and surgical technique (open approach) adversely affected LOS. Postoperative complications accounted for 40% of the significant individual predictors of LOS and included cardiac failure, deep vein thrombosis (DVT), intra-abdominal abscess, renal failure, and small bowel obstruction. During the course of the study, the percentage of colon resections performed laparoscopically increased from 4% to as high as 60% (Fig. 1). The median LOS was significantly TABLE 4 Odds Radios for Age as a Predictor of Prolonged Hospital Length of Stay Based on Differences in Patient Age Difference in age (years)
Odds ratio
⫹5 ⫹10 ⫹20 ⫹30 ⫹40
1.1 1.3 1.6 2.1 2.6
Note. A patient is 1.1 times more likely to have a prolonged LOS compared with a patient 5 years younger.
FIG. 1. Trend of cases performed laparoscopically over the study period. (Color version of figure is available online.)
longer for the first 225 colectomies studied (8 d) compared with the most recent 225 cases (6 d) (P ⬍ 0.0001). DISCUSSION
Colonic resections are common operations that are indicated for a wide range of malignant and benign pathologies. The morbidity rates of 35% to 40% reported in the literature can add significantly to the hospital LOS and the postoperative costs. Advanced age was the only demographic variable predictive of prolonged LOS. Other studies [6, 7] have also demonstrated advanced age to be a predictor of prolonged hospitalization. Many studies demonstrate advanced age to be associated with increases in postoperative morbidity and mortality for many types of surgical procedures [8 –12]; however, we did not specifically evaluate this relationship in our study. We showed that, as age increases, the likelihood of having a prolonged LOS increases. For two patients with ages 5 y apart, the older patient is 1.1 times more likely to have a prolonged LOS than the younger patients; for patients with ages 40 y apart, the ratio increases to 2.6 times (Table 4). We speculate that advanced age may lead to increased risk of postoperative morbidity, which in turn results in prolonged LOS. Of the preoperative medications considered, warfarin sodium was the only drug predictive of a prolonged hospital LOS. Several theories might explain this observation. The need for additional in-hospital time to allow a patient’s ProTime to return to a therapeutic range of anticoagulation after discontinuing the warfarin sodium perioperatively seems very likely. Additionally, this could also be related to bleeding complications at the time of surgery or in the postoperative period for patients who did not have their warfarin sodium fully reversed by the time of surgery due to the urgency of the operation. In addition, the reason that a patient needs warfarin
SCHMELZER ET AL.: FACTORS AFFECTING LENGTH OF STAY AFTER COLON RESECTION
sodium preoperatively (heart disease, thromboembolism, etc.) may add to the LOS. The use of chronic steroids, which has been associated with increased postoperative mortality [13], was associated with a longer LOS; however, multiple logistic regression analyses showed no predictive value for LOS in our study. Of the preoperative comorbidities, only alcoholism, end-stage renal disease, and illicit drug abuse were significant predictors of LOS. Alcoholism has been shown in other studies to result in higher risks of intensive care unit admissions postoperatively due to complications of alcohol withdrawal [14]. The same may apply to illicit drug use. There have been very few published reports evaluating illicit drug use and surgical outcomes. One study reported illicit drug use to be a predictor of significant postoperative morbidity [15]. End-stage renal disease is often reported as being a risk factor for increased postoperative morbidity and mortality, which is likely the reason for it to be predictive of prolonged LOS. The comorbidities often reported as being associated with increased morbidity and mortality (coronary artery disease, cerebrovascular accidents, hypertension, and morbid obesity) were not independently predictive of a prolonged hospital LOS. Several studies demonstrate higher morbidity and mortality associated with diabetes mellitus, cirrhosis, ischemic heart disease, pulmonary disease, and obesity; however, none addresses the impact on LOS [8, 13, 16 –18]. Operative variables and postoperative management are two areas often studied when examining impact on LOS. Often, the major focus is the type of postoperative analgesia (PCA, epidural, or oral narcotic), laparoscopic versus open techniques, and early enteral feeding. Our study did not examine the effect of early enteral feeding. Many of the surgeons in this study begin enteral feeding at initiation of clinical bowel function. Ileus, defined as the need for nasogastric tube decompression beyond postoperative day 5, was not a predictor of LOS in our study, so it could be inferred that timing of feeding did not play a role in LOS. Some publications have suggested that the use of an epidural for postoperative pain management results in a shorter hospital LOS [1, 5]; however, additional studies suggest epidural use may have benefits but does not affect the LOS [19, 20]. In our study, overall analgesic use was a significant predictor; however, there was no specific type of analgesic (route or drug) that predicted LOS. The benefits of laparoscopic versus open colectomy have been extensively examined over the past 10 y. There is increasing evidence showing that laparoscopic colectomy is associated with less morbidity, less postoperative pain, and shorter LOS [21–26]. We observed laparoscopic colectomy to be predictive of shorter hos-
199
pital LOS. The other operative variable that we found to be predictive of LOS was the need for a total colectomy (compared with segmental resections). This is not a surprise, since other studies have shown total abdominal colectomy to be associated with failed attempts at early enteral feeding, which can lead to longer postoperative LOS [27]. An ASA score ⱖ 3 was predictive of prolonged LOS. ASA score is used as a measure of preoperative risk for general anesthesia. There have been other scales created to attempt to predict postoperative morbidity and mortality, including the physiological and operative severity score for enumeration of mortality and morbidity (POSSUM), which is rarely used in the United States. Although this scale has been shown to be valid for predicting morbidity and mortality after operations, it has not been used in relation to hospital LOS [11, 28 –31]. However, in a study by Markus et al., POSSUM was found to be less predictive of postoperative morbidity than the surgeon’s “gut feeling” [32]. Consideration should be given to using this scale, which has been shown to correlate with ASA score, and determine if it is accurate in predicting prolonged hospital LOS. The only other category of variables that was predictive of prolonged LOS was postoperative morbidity. The individual variables in this group accounted for almost half of the significant individual predictors of prolonged LOS after colectomy. The morbidity rate for our study was 23%. Most of the complications were medical in nature, and only small bowel obstruction and intra-abdominal abscess were surgical complications. This observation is logical, and the significant association between the postoperative complications and prolonged LOS is expected. The occurrence of postoperative complications, especially those of a medical nature (e.g., myocardial infarction and respiratory failure), will likely lead to postoperative admissions to the intensive care unit and the need to implement additional therapeutic interventions such as endotracheal intubation, mechanical ventilation, vasomotor support, and dialysis. This will clearly impact the overall hospital LOS negatively. This study has certain limitations. The retrospective nature of the study makes it subject to unavoidable limitations in data collection. Due to the large number of patients reviewed, the information collected was limited to predetermined variables. We did not consider some variables such as preoperative albumin and timing to enteral feeding. Also, this study spans 8 y. During this time, many operative and postoperative strategies changed. The percentage of laparoscopic colectomies increased dramatically and the median postoperative LOS decreased. It is difficult to determine if the decrease in LOS was due to more cases being performed laparoscopically
200
JOURNAL OF SURGICAL RESEARCH: VOL. 146, NO. 2, MAY 15, 2008
or if the decrease was influenced more by pressure from the environment of cost minimization. This study, unlike most studies, examines the hospital LOS as the outcome variable. Previous reports focused on postoperative morbidity as the outcome variable and examined the predictors of postoperative complications. The lack of uniform standardized, criteria for the classification and severity of postoperative events is a limitation of most studies of surgical outcome. We, therefore, consider the hospital LOS to be representative of the quality and cost effectiveness of care and to provide a more clearly defined end point to surgical outcome. In summary, we identified the predictors of postoperative LOS to be a preoperative history of alcoholism, end-stage renal disease, or illicit drug use, patients on warfarin sodium, ASA score ⱖ 3, total colectomy, an open operation, and postoperative morbidity (including CVA, cardiac failure, DVT, respiratory failure, intraabdominal abscess, renal failure, urinary tract infection, or wound dehiscence). Awareness of these factors will help in implementing specific measures to improve outcome. Consideration should be given to preoperative optimization of high risk patients based on ASA score. If the colon resection can be adequately performed laparoscopically, it should be strongly considered. Postoperatively, efforts should be made to prevent medical complications (i.e., appropriate DVT prophylaxis, aggressive pulmonary toilet, etc).
11.
12.
13.
14.
15.
16.
17.
18.
19.
REFERENCES 1.
2.
3.
4. 5.
6. 7.
8.
Basse L, Thorbol JE, Lossl K, et al. Colonic surgery with accelerated rehabilitation or conventional care. Dis Colon Rectum 2004;47:271;Discussion 277. Basse L, Hjort Jakobsen D, Billesbolle P, et al. A clinical pathway to accelerate recovery after colonic resection. Ann Surg 2000;232:51. Bradshaw BG, Liu SS, Thirlby RC. Standardized perioperative care protocols and reduced length of stay after colon surgery. J Am Coll Surg 1998;186:501. Bufo AJ, Feldman S, Daniels GA, et al. Early postoperative feeding. Dis Colon Rectum 1994;37:1260. de Leon-Casasola OA, Parker BM, Lema MJ, et al. Epidural analgesia versus intravenous patient-controlled analgesia. Differences in the postoperative course of cancer patients. Reg Anesth 1994;19:307. Schoetz DJ, Jr., Bockler M, Rosenblatt MS, et al. “Ideal” length of stay after colectomy: Whose ideal? Dis Colon Rectum 1997;40:806. Tartter PI. Determinants of postoperative stay in patients with colorectal cancer. Implications for diagnostic-related groups. Dis Colon Rectum 1988;31:694. Alves A, Panis Y, Mathieu P, et al. Postoperative mortality and morbidity in French patients undergoing colorectal surgery: Results of a prospective multicenter study. Arch Surg 2005;140: 278; Discussion 284.
9.
Atkins BZ, Shah AS, Hutcheson KA, et al. Reducing hospital morbidity and mortality following esophagectomy. Ann Thorac Surg 2004;78:1170; Discussion 1170.
10.
de la Cruz Perez C, Estecha Foncea MA, Cruz Manas J, et al.
20.
21. 22.
23.
24. 25.
26.
27.
28.
Postoperative cardiac morbidity/mortality in high-risk elderly patients undergoing noncardiac surgery. Rev Esp Anestesiol Reanim 1999;46:4. Oomen JL, Engel AF, Cuesta MA. Mortality after acute surgery for complications of diverticular disease of the sigmoid colon is almost exclusively due to patient related factors. Colorectal Dis 2006;8:112–119. Sauvanet A, Mariette C, Thomas P, et al. Mortality and morbidity after resection for adenocarcinoma of the gastroesophageal junction: Predictive factors. J Am Coll Surg 2005;201:253. Margenthaler JA, Longo WE, Virgo KS, et al. Risk factors for adverse outcomes after the surgical treatment of appendicitis in adults. Ann Surg 2003;238:59. Maxson PM, Schultz KL, Berge KH, et al. Probable alcohol abuse or dependence: A risk factor for intensive care readmission in patients undergoing elective vascular and thoracic surgical procedures. Mayo Perioperative Outcomes Group. Mayo Clin Proc 1999;74:448. Misra P, Caldito GC, Kakkar AK, et al. Outcomes of coronary artery bypass grafting in patients with a history of illicit drug use. Am J Cardiol 2003;92:593. Pessaux P, Muscari F, Ouellet JF, et al. Risk factors for mortality and morbidity after elective sigmoid resection for diverticulitis: Prospective multicenter multivariate analysis of 582 patients. World J Surg 2004;28:92. Pedersen T. Complications and death following anaesthesia. A prospective study with special reference to the influence of patient-, anaesthesia-, and surgery-related risk factors. Dan Med Bull 1994;41:319. 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. Zutshi M, Delaney CP, Senagore AJ, et al. Randomized controlled trial comparing the controlled rehabilitation with early ambulation and diet pathway versus the controlled rehabilitation with early ambulation and diet with preemptive epidural anesthesia/analgesia after laparotomy and intestinal resection. Am J Surg 2005;189:268. Carli F, Trudel JL, Belliveau P. The effect of intraoperative thoracic epidural anesthesia and postoperative analgesia on bowel function after colorectal surgery: A prospective, randomized trial. Dis Colon Rectum 2001;44:1083. Ballantyne GH. Laparoscopic-assisted colorectal surgery: Review of results in 752 patients. Gastroenterologist 1995;3:75. Bardram L, Funch-Jensen PM, Jensen P, et al. Two days hospital stay after laparoscopic colon resection. Ugeskr Laeger 1996;158:5920. Chen HH, Wexner SD, Iroatulam AJ, et al. Laparoscopic colectomy compares favorably with colectomy by laparotomy for reduction of postoperative ileus. Dis Colon Rectum 2000;43:61. Falk PM, Beart RW, Jr., Wexner SD, et al. Laparoscopic colectomy: A critical appraisal. Dis Colon Rectum 1993;36:28. Hong D, Tabet J, Anvari M. Laparoscopic versus open resection for colorectal adenocarcinoma. Dis Colon Rectum 2001;44:10; Discussion 18. Veldkamp R, Kuhry E, Hop WC, et al. Laparoscopic surgery versus open surgery for colon cancer: Short-term outcomes of a randomized trial. Lancet Oncol 2005;6:477. Di Fronzo LA, Cymerman J, O’Connell TX. Factors affecting early postoperative feeding following elective open colon resection. Arch Surg 1999;134:941;Discussion 945. Ramkumar T, Ng V, Fowler L, et al. A comparison of POSSUM, P-POSSUM and colorectal POSSUM for the prediction of postoperative mortality in patients undergoing colorectal resection. Dis Colon Rectum 2006;49:330.
SCHMELZER ET AL.: FACTORS AFFECTING LENGTH OF STAY AFTER COLON RESECTION 29.
Al-Homoud S, Purkayastha S, Aziz O, et al. Evaluating operative risk in colorectal cancer surgery: ASA and POSSUM-based predictive models. Surg Oncol 2004;13:83. 30. Oomen JL, Engel AF, Cuesta MA. Outcome of elective primary surgery for diverticular disease of the sigmoid colon: A risk analysis based on the POSSUM scoring system. Colorectal Dis 2006;8:91.
31.
201
Tekkis PP, Kocher HM, Bentley AJ, et al. Operative mortality rates among surgeons: Comparison of POSSUM and p-POSSUM scoring systems in gastrointestinal surgery. Dis Colon Rectum 2000;43:1528;Discussion 1532. 32. Markus PM, Martell J, Leister I, et al. Predicting postoperative morbidity by clinical assessment. Br J Surg 2005;92:101.
Factors Affecting Length of Stay Following Colonic Resection Thomas M. Schmelzer, M.D., Gamal Mostafa, M.D.,1 Amy E. Lincourt, Ph.D., Steven M. Camp, M.D., Kent W. Kercher, M.D., Timothy S. Kuwada, M.D., and B. Todd Heniford, M.D. Department of Surgery, Carolinas Medical Center, Charlotte, North Carolina Submitted for publication February 13, 2007
Introduction. In-hospital length of stay (LOS) has become a valuable measure of outcomes following any operation, which also directly impacts cost. The aim of this study was to identify the factors that affect LOS after colonic resection. Materials and methods. A retrospective analysis was performed of adult patients who underwent colonic resection over an 8-y period at a tertiary institution. Data collected included demographics, American Society of Anesthesiologists (ASA) score, preoperative comorbidities and medications, operative management, postoperative morbidity and mortality, and LOS. Statistical analysis included descriptive statistics and multiple logistic regression to identify variables predictive of prolonged LOS. Results and discussion. A total of 899 consecutive patients were identified. One hundred eighty-seven resections were performed urgently, and 712 were elective. Two-hundred forty-five cases were performed laparoscopically. Complications occurred in 205 cases (23%), and there were 32 deaths (4%). The median LOS was 7 d. Logistic analysis showed 15 variables to be predictive of prolonged LOS. These included advanced age, warfarin sodium use, ASA score >3, alcoholism, chronic obstructive pulmonary disease, end-stage renal disease, illicit drug use, total colectomy (versus segmental), open resections (versus laparoscopic), and postoperative complications. In addition, the presence of at least one postoperative complication was predictive of prolonged LOS (P ⴝ 0.0002, OR 2.4 95% CI 1.5–3.8). Conclusions. ASA score and the incidence of postoperative complications are the only significant categories of variables that predict prolonged LOS after co-
1
To whom correspondence and reprint requests should be addressed at Department of Surgery/MEB 601, Carolinas Medical Center, 1000 Blythe Boulevard, Charlotte, NC 28203. E-mail: Gamal.Mostafa@ carolinashealthcare.org.
lectomy. Laparoscopic approach and the extent of the resection are predictive as well. © 2008 Elsevier Inc. All rights reserved.
Key Words: adult; colectomy; colon resection; human; laparoscopic colectomy; length of stay; morbidity; multiple linear regression; risk factors.
INTRODUCTION
Hospital length of stay (LOS) after major surgery is becoming an increasingly important measurement of surgical outcome. The influence of the insurance industry has led to a shift toward more cost-effective management of postoperative patients, which includes early release from the hospital. Surgeons are evaluated based on their ability to deliver quality care while minimizing unnecessary costs. The impact of this shift can be seen in colonic resections. This relatively common operation was frequently associated with a 6 to 12 d postoperative hospital LOS [1]. Strict postoperative protocols have been developed to “fast track” the postoperative course with the goal LOS being 2 d [2]. The methods used to decrease the recovery time include using laparoscopic techniques, early postoperative feeding, epidurals for pain management, and early mobilization [1–5]. These methods attempt to address the common limiting factors for most patients, which are pain and bowel function; however, they do not address the issues of increased LOS due to postoperative morbidity. This study examines the potential impact of various demographics, preoperative morbidities, operative details, and postoperative morbidities on hospital LOS after colonic resection. Identifying risk factors for extended LOS can facilitate modification to the preoperative preparation and postoperative management of patients undergoing colonic resections, thereby leading to more cost-effective delivery of surgical care.
195
0022-4804/08 $34.00 © 2008 Elsevier Inc. All rights reserved.
196
JOURNAL OF SURGICAL RESEARCH: VOL. 146, NO. 2, MAY 15, 2008
METHODS After obtaining Institutional Review Board approval, a retrospective review of the medical records of adult patients (age ⬎ 18 y) undergoing a colonic resection was conducted at a single tertiary institution over an 8-y period (January 1998 to December 2005). Records were identified using ICD-9 codes for colonic resection. Trauma patients were excluded from the study. Six categories of variables were used in data collection, including demographics, preoperative comorbidities, preoperative medications, American Society of Anesthesiologist (ASA) score, operative details, and postoperative morbidity/mortality. All independent variables in these categories are shown in Table 1. For this study, urgent operations were defined as operations performed within 24 h of admission or the time of surgical consultation. Comorbidity was defined as illnesses or conditions that existed before the operation and not related to or caused by the colonic pathology. A right colectomy was defined as a resection of the colon proximal to the middle colic artery; a left colectomy as a resection of the colon distal to the middle colic artery; a total colectomy as a resection of the full length of the colon from the terminal ileum to the rectum. Analgesia used referred to the amount of postoperative narcotic used by patients (measured in morphine equivalents). A wound infection was defined as erythema of the skin incision or culture-proved wound drainage requiring antibiotics or local wound care intervention. Wound dehiscence was defined as disrup-
TABLE 1 Data Evaluated in 899 Patients Undergoing Colectomy Demographics Age BMI Comorbidities Alcoholism Asthma CHF Cirrhosis/liver disease COPD Coronary artery disease Cerebrovascular accident Diabetes Preoperative medications Antiplatelet (aspirin or clopidogrel) Warfarin sodium ASA Score Operative Information Analgesia Blood transfusion Estimated blood loss Operative time Pathologic diagnosis Postoperative morbidity Anastomotic leak Atelectasis Bleeding Cardiac events (MI, failure, dysrhythmia) Cerebrovascular accident Deep vein thrombosis Ileus Intra-abdominal abscess Pneumonia
Race Sex End stage renal disease HIV Hypertension Illicit drug use Morbid obesity Previous intra-abdominal surgery Renal insufficiency (Cr. ⬎ 1.3) Smoking Steroids
Procedure (left/right/segmental/ total colectomy) Surgical technique (laparoscopic versus open) Urgency (urgent or elective) Pulmonary embolism Renal failure Respiratory failure Sepsis Small bowel obstruction Urinary tract infection Wound dehiscence Wound infection
BMI ⫽ body mass index; CHF ⫽ congestive heart failure; COPD ⫽ chronic obstructive pulmonary disease; MI ⫽ myocardial infarction.
tion of the fascial closure resulting in exposed viscera and requiring surgical intervention. An abdominal computed tomography scan was used to diagnose intra-abdominal abscess. A computed tomography scan with oral or rectal contrast was used to diagnose or exclude anastomotic leak. Clinically significant ileus was defined as requiring nasogastric tube decompression beyond postoperative day 5. Atelectasis was recorded based on radiographic reports from chest X-rays. Bleeding was defined as a drop in hemoglobin of 2 g/dL or requiring the transfusion of two or more units of packed red blood cells within the first 48 h after colectomy. Renal failure was defined as an increase in serum creatinine by a factor of 3, a decrease in glomerular filtration by 75%, or urine output less than 0.3 mL/kg/h for a period of 24 h. Respiratory failure was defined as requiring intubation and mechanical ventilation. Cardiac events were determined by cardiac enzyme values and electrocardiogram findings. Small bowel obstruction was defined by abdominal distention with or without obstipation and radiographic findings consistent with obstruction. The primary end point for this study was postoperative hospital LOS measured in days. Data were analyzed using standard statistical methods. Descriptive statistics including means and standard deviations or counts and percentages were used to describe the study population on all variables. Differences between the prolonged LOS and nonprolonged LOS groups were calculated using Wilcoxon for continuous variables and -square or Fisher’s exact for categorical variables. Significant variables were then selected using backward elimination to determine significant independent predictors of prolonged LOS using the Wald -square test logistic regression analysis. A prolonged LOS was defined as greater than the median LOS for the group, since the LOS was not normally distributed. Comparisons were also made between the median LOS for the first 25% of the observations and the most recent 25% of observations to determine differences over the time of the study. A P-value ⬍0.05 was considered significant for all determinations. SAS System software, version 8.02 (SAS Institute, Cary, NC) was used for statistical analyses.
RESULTS
During the study period, 899 colonic resections were identified. There were 449 males and 450 female patients. The average age of the patients was 61.8 y (range 19 to 92 years). One-hundred eighty-seven resections (21%) were performed urgently, and 712 (79%) were elective. Two-hundred forty-five cases (27%) were completed laparoscopically, 45 (5%) were started laparoscopically and converted to open, and 609 (68%) were open colon resections. Cancer was the most common indication for colectomy (52%), followed by diverticular disease (34%), and colitis (9%). Four-hundred thirtythree (48%) of the resections were right colectomies, 407 (45%) were left colectomies, and there were 59 (7%) total colectomies. Complications were reported in 205 cases (23%), and there were 32 deaths (4%). The median LOS was 7 d (range 1 to 138). Descriptive statistics for the prolonged versus nonprolonged LOS groups can be seen in Table 2. Twenty-four variables were statistically different between the prolonged LOS and nonprolonged LOS groups. Logistic regression analysis of these variables showed 15 variables to be predictive of prolonged LOS (Table 3). The presence of at least one postoperative complication was also a significant predictor of prolonged LOS (P ⫽ 0.0002, OR ⫽ 2.4 95% CI 1.5–3.8).
SCHMELZER ET AL.: FACTORS AFFECTING LENGTH OF STAY AFTER COLON RESECTION
197
TABLE 2 Descriptive Statistics for All Variables Based on Prolonged versus Nonprolonged Hospital Length of Stay Nonprolonged LOS (N ⫽ 375) Demographics Mean age (years) BMI (kg/m 2) Race (Caucasian/African American/Hispanic) Sex (male/female) Comorbidities (overall) Alcoholism Asthma CHF Cirrhosis/liver disease COPD Coronary artery disease CVA Diabetes End stage renal disease HIV Hypertension Illicit drug use Morbid obesity Renal insufficiency (Cr. ⬎ 1.3) Smoking Preoperative medications Antiplatelet (aspirin or clopidogrel bisulfate) Warfarin sodium Steroids ASA score (⬍3/ⱖ3) Operative information Analgesia (epidural) Blood transfusion Estimated blood loss (mL) Operative time (min) Pathologic diagnosis (malignant/benign) Procedure (left/right/total colectomy) Surgical technique (laparoscopic versus open) Urgency (urgent or elective) Postoperative morbidity (at least one) Anastomotic leak Atelectasis Bleeding Cardiac events (MI, failure, dysrhythmia) CVA Deep vein thrombosis Ileus Intra-abdominal abscess Pneumonia Pulmonary embolism Renal failure Respiratory failure Sepsis Small bowel obstruction Urinary tract infection Wound dehiscence Wound infection
Prolonged LOS (N ⫽ 524)
P-value
57 28.1 298/66/5 178/197 204 6 21 7 1 14 32 5 97 2 0 136 2 13 11 46
62 26.8 268/133/6 271/253 302 24 7 14 3 57 93 15 162 20 3 233 11 10 27 57
⬍0.0001 0.02 0.09 0.2 0.9 0.05 ⬍0.0001 0.4 0.6 ⬍0.0001 ⬍0.0001 0.2 0.1 0.001 0.3 0.01 0.1 0.1 0.2 0.5
5 4 11 266/109
10 19 13 231/293
0.1 ⬍0.0001 0.5 ⬍0.0001
14 7 156 177 167/208 175/186/14 208/167 39/336 61 1 23 4 0 0 0 10 0 1 2 0 1 3 3 2 0 14
29 39 208 177 299/225 232/247/45 95/429 147/377 144 4 34 6 12 1 12 37 6 6 3 6 14 10 15 16 3 17
0.2 ⬍0.0001 0.8 0.9 0.0002 0.02 ⬍0.0001 ⬍0.0001 ⬍0.0001 0.2 1 1 0.002 1 0.002 ⬍0.004 0.05 0.2 1 0.04 0.001 0.2 0.03 0.005 0.2 1
BMI ⫽ body mass index; CHF ⫽ congestive heart failure; COPD ⫽ chronic obstructive pulmonary disease; CVA ⫽ cerebrovascular accident; MI ⫽ myocardial infarction.
The only preoperative category of variables that was predictive of prolonged LOS was ASA score ⱖ 3 (P ⫽ 0.003, OR ⫽ 3.5; 95% CI 1.6 –9.3). Demographics, comorbidities, preoperative medications, and operative
variables were not categorically predictive of prolonged LOS. Age was a significant predictor of prolonged LOS, and patients with advanced age were increasingly more likely to have a prolonged LOS (Table 4). Warfarin
198
JOURNAL OF SURGICAL RESEARCH: VOL. 146, NO. 2, MAY 15, 2008
TABLE 3 Significant Predictors of Hospital Length of Stay Identified by Logistic Regression P-value Demographics Age Comorbidities (at least one) Alcoholism COPD ESRD Illicit drug use Preoperative medications Warfarin sodium ASA grade (ⱖ3) Operative information Procedure (total colectomy) Surgical technique (open approach) Postoperative morbidity Cardiac failure DVT Intra-abdominal abscess Renal failure Respiratory failure Small bowel obstruction
Odds ratio
95% CI
⬍0.0001 0.3* 0.01 0.004 0.03 ⬍0.0001
N/A
N/A
2.2 3.1 6.2 10.0
1.2–4.1 1.4–6.7 1.2–33.3 4.1–24.4
0.04 0.003
6.0 3.5
1.1–26.7 1.6–9.3
0.02
4.2
1.4–13.3
⬍0.0001 0.0002 0.006 ⬍0.0001 0.002 0.02 0.02 0.04
5.3 2.4 2.8 2.6 2.9 3.3 3.7 6.0
3.4–8.5 1.5–3.8 1.3–6.5 1.6–4.1 1.5–6.0 1.3–12.5 1.8–9.3 1.1–33.3
CI ⫽ confidence interval; COPD ⫽ chronic obstructive pulmonary disease; ESRD ⫽ end-stage renal disease. * Nonsignificant value.
sodium was the only preoperative medication predictive of a prolonged LOS (P ⫽ 0.04). Comorbidities serving as significant predictors of LOS were alcoholism, chronic obstructive pulmonary disease, end-stage renal disease, and illicit drug abuse. The operative variables total colectomy and surgical technique (open approach) adversely affected LOS. Postoperative complications accounted for 40% of the significant individual predictors of LOS and included cardiac failure, deep vein thrombosis (DVT), intra-abdominal abscess, renal failure, and small bowel obstruction. During the course of the study, the percentage of colon resections performed laparoscopically increased from 4% to as high as 60% (Fig. 1). The median LOS was significantly TABLE 4 Odds Radios for Age as a Predictor of Prolonged Hospital Length of Stay Based on Differences in Patient Age Difference in age (years)
Odds ratio
⫹5 ⫹10 ⫹20 ⫹30 ⫹40
1.1 1.3 1.6 2.1 2.6
Note. A patient is 1.1 times more likely to have a prolonged LOS compared with a patient 5 years younger.
FIG. 1. Trend of cases performed laparoscopically over the study period. (Color version of figure is available online.)
longer for the first 225 colectomies studied (8 d) compared with the most recent 225 cases (6 d) (P ⬍ 0.0001). DISCUSSION
Colonic resections are common operations that are indicated for a wide range of malignant and benign pathologies. The morbidity rates of 35% to 40% reported in the literature can add significantly to the hospital LOS and the postoperative costs. Advanced age was the only demographic variable predictive of prolonged LOS. Other studies [6, 7] have also demonstrated advanced age to be a predictor of prolonged hospitalization. Many studies demonstrate advanced age to be associated with increases in postoperative morbidity and mortality for many types of surgical procedures [8 –12]; however, we did not specifically evaluate this relationship in our study. We showed that, as age increases, the likelihood of having a prolonged LOS increases. For two patients with ages 5 y apart, the older patient is 1.1 times more likely to have a prolonged LOS than the younger patients; for patients with ages 40 y apart, the ratio increases to 2.6 times (Table 4). We speculate that advanced age may lead to increased risk of postoperative morbidity, which in turn results in prolonged LOS. Of the preoperative medications considered, warfarin sodium was the only drug predictive of a prolonged hospital LOS. Several theories might explain this observation. The need for additional in-hospital time to allow a patient’s ProTime to return to a therapeutic range of anticoagulation after discontinuing the warfarin sodium perioperatively seems very likely. Additionally, this could also be related to bleeding complications at the time of surgery or in the postoperative period for patients who did not have their warfarin sodium fully reversed by the time of surgery due to the urgency of the operation. In addition, the reason that a patient needs warfarin
SCHMELZER ET AL.: FACTORS AFFECTING LENGTH OF STAY AFTER COLON RESECTION
sodium preoperatively (heart disease, thromboembolism, etc.) may add to the LOS. The use of chronic steroids, which has been associated with increased postoperative mortality [13], was associated with a longer LOS; however, multiple logistic regression analyses showed no predictive value for LOS in our study. Of the preoperative comorbidities, only alcoholism, end-stage renal disease, and illicit drug abuse were significant predictors of LOS. Alcoholism has been shown in other studies to result in higher risks of intensive care unit admissions postoperatively due to complications of alcohol withdrawal [14]. The same may apply to illicit drug use. There have been very few published reports evaluating illicit drug use and surgical outcomes. One study reported illicit drug use to be a predictor of significant postoperative morbidity [15]. End-stage renal disease is often reported as being a risk factor for increased postoperative morbidity and mortality, which is likely the reason for it to be predictive of prolonged LOS. The comorbidities often reported as being associated with increased morbidity and mortality (coronary artery disease, cerebrovascular accidents, hypertension, and morbid obesity) were not independently predictive of a prolonged hospital LOS. Several studies demonstrate higher morbidity and mortality associated with diabetes mellitus, cirrhosis, ischemic heart disease, pulmonary disease, and obesity; however, none addresses the impact on LOS [8, 13, 16 –18]. Operative variables and postoperative management are two areas often studied when examining impact on LOS. Often, the major focus is the type of postoperative analgesia (PCA, epidural, or oral narcotic), laparoscopic versus open techniques, and early enteral feeding. Our study did not examine the effect of early enteral feeding. Many of the surgeons in this study begin enteral feeding at initiation of clinical bowel function. Ileus, defined as the need for nasogastric tube decompression beyond postoperative day 5, was not a predictor of LOS in our study, so it could be inferred that timing of feeding did not play a role in LOS. Some publications have suggested that the use of an epidural for postoperative pain management results in a shorter hospital LOS [1, 5]; however, additional studies suggest epidural use may have benefits but does not affect the LOS [19, 20]. In our study, overall analgesic use was a significant predictor; however, there was no specific type of analgesic (route or drug) that predicted LOS. The benefits of laparoscopic versus open colectomy have been extensively examined over the past 10 y. There is increasing evidence showing that laparoscopic colectomy is associated with less morbidity, less postoperative pain, and shorter LOS [21–26]. We observed laparoscopic colectomy to be predictive of shorter hos-
199
pital LOS. The other operative variable that we found to be predictive of LOS was the need for a total colectomy (compared with segmental resections). This is not a surprise, since other studies have shown total abdominal colectomy to be associated with failed attempts at early enteral feeding, which can lead to longer postoperative LOS [27]. An ASA score ⱖ 3 was predictive of prolonged LOS. ASA score is used as a measure of preoperative risk for general anesthesia. There have been other scales created to attempt to predict postoperative morbidity and mortality, including the physiological and operative severity score for enumeration of mortality and morbidity (POSSUM), which is rarely used in the United States. Although this scale has been shown to be valid for predicting morbidity and mortality after operations, it has not been used in relation to hospital LOS [11, 28 –31]. However, in a study by Markus et al., POSSUM was found to be less predictive of postoperative morbidity than the surgeon’s “gut feeling” [32]. Consideration should be given to using this scale, which has been shown to correlate with ASA score, and determine if it is accurate in predicting prolonged hospital LOS. The only other category of variables that was predictive of prolonged LOS was postoperative morbidity. The individual variables in this group accounted for almost half of the significant individual predictors of prolonged LOS after colectomy. The morbidity rate for our study was 23%. Most of the complications were medical in nature, and only small bowel obstruction and intra-abdominal abscess were surgical complications. This observation is logical, and the significant association between the postoperative complications and prolonged LOS is expected. The occurrence of postoperative complications, especially those of a medical nature (e.g., myocardial infarction and respiratory failure), will likely lead to postoperative admissions to the intensive care unit and the need to implement additional therapeutic interventions such as endotracheal intubation, mechanical ventilation, vasomotor support, and dialysis. This will clearly impact the overall hospital LOS negatively. This study has certain limitations. The retrospective nature of the study makes it subject to unavoidable limitations in data collection. Due to the large number of patients reviewed, the information collected was limited to predetermined variables. We did not consider some variables such as preoperative albumin and timing to enteral feeding. Also, this study spans 8 y. During this time, many operative and postoperative strategies changed. The percentage of laparoscopic colectomies increased dramatically and the median postoperative LOS decreased. It is difficult to determine if the decrease in LOS was due to more cases being performed laparoscopically
200
JOURNAL OF SURGICAL RESEARCH: VOL. 146, NO. 2, MAY 15, 2008
or if the decrease was influenced more by pressure from the environment of cost minimization. This study, unlike most studies, examines the hospital LOS as the outcome variable. Previous reports focused on postoperative morbidity as the outcome variable and examined the predictors of postoperative complications. The lack of uniform standardized, criteria for the classification and severity of postoperative events is a limitation of most studies of surgical outcome. We, therefore, consider the hospital LOS to be representative of the quality and cost effectiveness of care and to provide a more clearly defined end point to surgical outcome. In summary, we identified the predictors of postoperative LOS to be a preoperative history of alcoholism, end-stage renal disease, or illicit drug use, patients on warfarin sodium, ASA score ⱖ 3, total colectomy, an open operation, and postoperative morbidity (including CVA, cardiac failure, DVT, respiratory failure, intraabdominal abscess, renal failure, urinary tract infection, or wound dehiscence). Awareness of these factors will help in implementing specific measures to improve outcome. Consideration should be given to preoperative optimization of high risk patients based on ASA score. If the colon resection can be adequately performed laparoscopically, it should be strongly considered. Postoperatively, efforts should be made to prevent medical complications (i.e., appropriate DVT prophylaxis, aggressive pulmonary toilet, etc).
11.
12.
13.
14.
15.
16.
17.
18.
19.
REFERENCES 1.
2.
3.
4. 5.
6. 7.
8.
Basse L, Thorbol JE, Lossl K, et al. Colonic surgery with accelerated rehabilitation or conventional care. Dis Colon Rectum 2004;47:271;Discussion 277. Basse L, Hjort Jakobsen D, Billesbolle P, et al. A clinical pathway to accelerate recovery after colonic resection. Ann Surg 2000;232:51. Bradshaw BG, Liu SS, Thirlby RC. Standardized perioperative care protocols and reduced length of stay after colon surgery. J Am Coll Surg 1998;186:501. Bufo AJ, Feldman S, Daniels GA, et al. Early postoperative feeding. Dis Colon Rectum 1994;37:1260. de Leon-Casasola OA, Parker BM, Lema MJ, et al. Epidural analgesia versus intravenous patient-controlled analgesia. Differences in the postoperative course of cancer patients. Reg Anesth 1994;19:307. Schoetz DJ, Jr., Bockler M, Rosenblatt MS, et al. “Ideal” length of stay after colectomy: Whose ideal? Dis Colon Rectum 1997;40:806. Tartter PI. Determinants of postoperative stay in patients with colorectal cancer. Implications for diagnostic-related groups. Dis Colon Rectum 1988;31:694. Alves A, Panis Y, Mathieu P, et al. Postoperative mortality and morbidity in French patients undergoing colorectal surgery: Results of a prospective multicenter study. Arch Surg 2005;140: 278; Discussion 284.
9.
Atkins BZ, Shah AS, Hutcheson KA, et al. Reducing hospital morbidity and mortality following esophagectomy. Ann Thorac Surg 2004;78:1170; Discussion 1170.
10.
de la Cruz Perez C, Estecha Foncea MA, Cruz Manas J, et al.
20.
21. 22.
23.
24. 25.
26.
27.
28.
Postoperative cardiac morbidity/mortality in high-risk elderly patients undergoing noncardiac surgery. Rev Esp Anestesiol Reanim 1999;46:4. Oomen JL, Engel AF, Cuesta MA. Mortality after acute surgery for complications of diverticular disease of the sigmoid colon is almost exclusively due to patient related factors. Colorectal Dis 2006;8:112–119. Sauvanet A, Mariette C, Thomas P, et al. Mortality and morbidity after resection for adenocarcinoma of the gastroesophageal junction: Predictive factors. J Am Coll Surg 2005;201:253. Margenthaler JA, Longo WE, Virgo KS, et al. Risk factors for adverse outcomes after the surgical treatment of appendicitis in adults. Ann Surg 2003;238:59. Maxson PM, Schultz KL, Berge KH, et al. Probable alcohol abuse or dependence: A risk factor for intensive care readmission in patients undergoing elective vascular and thoracic surgical procedures. Mayo Perioperative Outcomes Group. Mayo Clin Proc 1999;74:448. Misra P, Caldito GC, Kakkar AK, et al. Outcomes of coronary artery bypass grafting in patients with a history of illicit drug use. Am J Cardiol 2003;92:593. Pessaux P, Muscari F, Ouellet JF, et al. Risk factors for mortality and morbidity after elective sigmoid resection for diverticulitis: Prospective multicenter multivariate analysis of 582 patients. World J Surg 2004;28:92. Pedersen T. Complications and death following anaesthesia. A prospective study with special reference to the influence of patient-, anaesthesia-, and surgery-related risk factors. Dan Med Bull 1994;41:319. 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. Zutshi M, Delaney CP, Senagore AJ, et al. Randomized controlled trial comparing the controlled rehabilitation with early ambulation and diet pathway versus the controlled rehabilitation with early ambulation and diet with preemptive epidural anesthesia/analgesia after laparotomy and intestinal resection. Am J Surg 2005;189:268. Carli F, Trudel JL, Belliveau P. The effect of intraoperative thoracic epidural anesthesia and postoperative analgesia on bowel function after colorectal surgery: A prospective, randomized trial. Dis Colon Rectum 2001;44:1083. Ballantyne GH. Laparoscopic-assisted colorectal surgery: Review of results in 752 patients. Gastroenterologist 1995;3:75. Bardram L, Funch-Jensen PM, Jensen P, et al. Two days hospital stay after laparoscopic colon resection. Ugeskr Laeger 1996;158:5920. Chen HH, Wexner SD, Iroatulam AJ, et al. Laparoscopic colectomy compares favorably with colectomy by laparotomy for reduction of postoperative ileus. Dis Colon Rectum 2000;43:61. Falk PM, Beart RW, Jr., Wexner SD, et al. Laparoscopic colectomy: A critical appraisal. Dis Colon Rectum 1993;36:28. Hong D, Tabet J, Anvari M. Laparoscopic versus open resection for colorectal adenocarcinoma. Dis Colon Rectum 2001;44:10; Discussion 18. Veldkamp R, Kuhry E, Hop WC, et al. Laparoscopic surgery versus open surgery for colon cancer: Short-term outcomes of a randomized trial. Lancet Oncol 2005;6:477. Di Fronzo LA, Cymerman J, O’Connell TX. Factors affecting early postoperative feeding following elective open colon resection. Arch Surg 1999;134:941;Discussion 945. Ramkumar T, Ng V, Fowler L, et al. A comparison of POSSUM, P-POSSUM and colorectal POSSUM for the prediction of postoperative mortality in patients undergoing colorectal resection. Dis Colon Rectum 2006;49:330.
SCHMELZER ET AL.: FACTORS AFFECTING LENGTH OF STAY AFTER COLON RESECTION 29.
Al-Homoud S, Purkayastha S, Aziz O, et al. Evaluating operative risk in colorectal cancer surgery: ASA and POSSUM-based predictive models. Surg Oncol 2004;13:83. 30. Oomen JL, Engel AF, Cuesta MA. Outcome of elective primary surgery for diverticular disease of the sigmoid colon: A risk analysis based on the POSSUM scoring system. Colorectal Dis 2006;8:91.
31.
201
Tekkis PP, Kocher HM, Bentley AJ, et al. Operative mortality rates among surgeons: Comparison of POSSUM and p-POSSUM scoring systems in gastrointestinal surgery. Dis Colon Rectum 2000;43:1528;Discussion 1532. 32. Markus PM, Martell J, Leister I, et al. Predicting postoperative morbidity by clinical assessment. Br J Surg 2005;92:101.