Surg Endosc (1997) 11: 272276

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Cost analysis of diagnostic laparoscopy vs laparotomy in the evaluation of penetrating abdominal trauma
J. M. Marks,1 D. F. Youngelman,2 T. Berk1
1

Department of Surgery, The Mount Sinai Medical Center, School of Medicine, Case Western Reserve University, One Mount Sinai Drive, Cleveland, OH 44106, USA 2 Department of Surgery, Harry S. Truman Memorial Veterans Hospital, 800 Hospital Drive, University of Missouri-Columbia, Columbia, MO 65201, USA Received: 11 March 1996/Accepted: 5 July 1996

Abstract Background: Diagnostic laparoscopy for the evaluation of injuries in patients with penetrating abdominal trauma has been shown to decrease the morbidity and mortality associated with mandatory laparotomy. The overall impact on patient care and hospital costs has not been thoroughly investigated. The goal of this study was to determine the economic impact of laparoscopy as a diagnostic tool in the management of patients following penetrating trauma to the abdomen or flank. Methods: Retrospective chart review of all hemodynamically stable patients with penetrating trauma to the abdomen or flank, but without other injuries requiring emergent intervention, admitted to a level I trauma center between January 1, 1992, and September 30, 1994. Those patients who underwent either laparoscopy (DL) or laparotomy (NL) or both (CONV) and who had no intraabdominal organ injuries requiring surgical therapeutic intervention were included in the study. Age, operative time, operative findings, length of hospitalization, Injury Severity Score (ISS), variable costs, and total costs were recorded for each patient. Results: Fourteen patients underwent negative/nontherapeutic laparoscopy (DL), 19 patients underwent negative/nontherapeutic laparotomy (NL), and four patients underwent both laparoscopy and laparotomy, a conversion procedure (CONV). There was no significant difference in age, operative times, or ISS between the DL and NL groups. Mean ISS of CONV patients was significantly greater than that of DL patients, 5.75 1.97 vs 2.43 0.63 (p < 0.05). Mean operative time for CONV patients was also significantly greater than both DL and NL patients, 106.5 17.00 min vs 66.1 6.55 and 47.3 7.50 min, respectively (p < 0.05). The mean length of stay was significantly shorter in

the DL group as compared to the NL or CONV groups, 1.43 0.20 vs 4.26 0.31 and 5.0 0.82 (p < 0.0001). The variable costs for the DL group were significantly lower than those incurred by patients in the NL and CONV groups, $2,917 175 vs $3,384 102 and $3,774 286, (p < 0.05). Variable costs were not significantly different between the NL and CONV groups. Total costs were also significantly lower in the DL group when compared to NL and CONV, $5,427 394 vs $7,026 251 and $7,855 750 (p < 0.005), but again, they were not statistically different between the NL and CONV groups. The overall total costs for laparoscopy, including the costs incurred by conversion patients, was significantly less than the total costs for laparotomy patients, $5,664 394 vs $7,028.47 250 (p < 0.005). This resulted in an overall savings of $1,059.44 per laparoscopy performed. The overall negative/nontherapeutic laparotomy rate during this study was 19.1%, which was significantly lower than the negative or nontherapeutic exploration rate during the time period prior to the use of laparoscopy (p < 0.01, z 2.550). Conclusion: Variable and total costs and length of stay were significantly lower in our population of patients who underwent DL as compared to NL. The rate of negative or nontherapeutic laparotomy was also significantly reduced when compared to the rate identified during the era prior to the use of laparoscopy. Laparoscopy resulted in an overall savings of $1,059 per laparoscopy performed when compared to laparotomy. Key words: Laparoscopy Penetrating abdominal trauma Cost effectiveness

Correspondence to: J. M. Marks Presented at the 5th World Congress of Endoscopic Surgery of the Society of American Gastrointestinal Endoscopic Surgeons (SAGES) Scientific Session, Philadelphia, Pennsylvania, USA, 15 March 1996

The use of diagnostic laparoscopy for the evaluation of penetrating abdominal trauma is gaining increasing acceptance. The high rates of negative and nontherapeutic laparotomy and their inherent complications are added incentive

273 Table 1. Patients admitted with penetrating trauma to the abdomen or flank between January 1, 1992, and September 30, 1994 Gunshot/shotgun victims (n 81) Exploratory laparotomy Negative/nontherapeutic Therapeutic Diagnostic laparoscopy Negative/nontherapeutic Therapeutic Conversion (laparoscopy/laparotomy) Negative/nontherapeutic Therapeutic 8 69 3 0 1 0 Stab wound victims (n 50) 11 24 11 1 2 1
a

Table 2. Comparison of the negative/nontherapeutic exploration rates for patients with penetrating trauma to the abdomen or flank Group I 19891991 (laparotomy only) Gunshot/shotgun victims Stab wound victims Total 28.2% (20/71) 42.6% (20/47) 33.9% (40/118) Group II 19921994 (laparotomy only) 11.5% (9/78) 35.1% (13/37) 19.1%a (22/115) Group III 19921994 (laparoscopy or laparotomy) 14.8% (12/81) 48.0% (24/50) 27.5% (36/131)

Group I vs group II (p < 0.01, z 2.550)

for the increasing use of this modality. Laparoscopy has been shown to be useful in excluding or confirming peritoneal penetration and therefore in preventing unnecessary laparotomy [25, 7, 8, 11, 15, 17]. Morbidity and mortality rates have also been shown to be significantly lower in patients undergoing laparoscopy as compared to celiotomy [2, 4, 17]. Length of hospital stay has also been proven to be significantly shorter following laparoscopic evaluation as compared to formal laparotomy [4, 7]. This retrospective study was undertaken to compare the costs incurred and the overall economic impact of laparoscopy as a diagnostic tool in the management of patients with penetrating trauma to the abdomen and flank. Materials and methods
The charts were reviewed of all patients admitted (to the Mt. Sinai Medical Center, Cleveland, Ohio, a level I trauma center) with penetrating trauma to the abdomen or flank between January 1, 1992, and September 30, 1994. These patients were all evaluated by the trauma team, which included a chief surgical resident and the surgical attending on call. Included in this study were all patients who underwent diagnostic nontherapeutic laparoscopy (DL) or negative or nontherapeutic laparotomy (NL). Conversion patients (CONV) were those patients who underwent both laparoscopy and laparotomy. Negative laparoscopy or laparotomy was defined as an exploration without evidence of peritoneal penetration or intraabdominal organ injury. Nontherapeutic laparoscopy or laparotomy was defined as an exploration identifying peritoneal penetration and/or intraabdominal organ injury which did not demand surgical therapeutic intervention. Hemodynamically stable patients (SBP > 90 mmHg, HR < 110) were candidates for diagnostic laparoscopy if they had (1) stab wounds to the abdomen or flank with presumptive peritoneal penetration based on omental evisceration, positive diagnostic peritoneal lavage or positive wound exploration, or (2) tangential gunshot wounds without obvious peritoneal penetration. Laparoscopy was not considered in patients who required other emergent procedures such as exploration for peripheral vascular injury or those who were not hemodynamically stable. All procedures were performed in the operating room under general anesthesia and all patients consented to possible conversion to laparotomy. After induction of general anesthesia, a Foley catheter and orogastric or nasogastric tube were placed in each patient. An umbilical nondisposable trocar (Stortz, Charlton, MA) was placed by the Hasson technique and the abdomen was insufflated with CO2 to a pressure of 15 mmHg. A 0, 10-mm laparoscope was used initially in all patients. A 30 laparoscope was used as needed for evaluation of the diaphragm and upper abdominal organs. Additional 5-mm ports were placed under direct vision as necessary for manipulation of the bowel. Nondisposable, noncrushing bowel clamps (Olympus, Melville, NY; Marlow, Willoughby, OH) were used routinely to decrease the risk of bowel injury and to lower the variable costs incurred by laparoscopy. All quadrants were carefully inspected and the small bowel and colon were examined completely. All exploratory laparotomies were performed in standard fashion through a midline inci-

sion under general anesthesia. Variable and total costs were determined for each patient by the hospital billing office. All costs were adjusted for inflation to 1994 rates. The overall economic impact of laparoscopy was calculated by averaging the additional expense incurred by the conversion group to the DL group (overall laparoscopy costs DL costs + [CONV costs NL costs] / of DL + CONV patients). Statistical analysis was done by Students paired t-test, or z-test analysis.

Results Between January 1, 1992, and September 30, 1994, 131 patients were admitted to the trauma service with penetrating injuries to the flank or abdomen. There were 81 patients with gunshot wounds and 50 patients with stab wounds (Table 1). Ninety-five percent of patients with gunshot wounds underwent laparotomy (77/81). DL was performed on 3.7% of gunshot victims (3/81), and CONV on 1.2% of these patients (1/81). Seventy percent of patients with stab wounds underwent laparotomy (35/50), 24% underwent DL (12/50), and 6% had CONV (3/50). The combined negative or nontherapeutic laparotomy/laparoscopy rate was 27.5% (GSW 14.8%, STAB 48.0%), which was comparable to the negative or nontherapeutic laparotomy rate of 33.9% (p > 0.10, as determined by z-test) for a similar time period between March 1, 1989 and December 31, 1991, prior to the use of laparoscopy. The overall negative/nontherapeutic laparotomy rate was 19.1%, which was significantly lower than the negative or nontherapeutic exploration rate during the time period prior to the use of laparoscopy (p < 0.01, z 2.550) (Table 2). Four patients with gunshot wounds were evaluated by DL (Table 3). All four were negative and one was converted to laparotomy, which confirmed the negative findings at laparoscopy. Fifteen patients underwent diagnostic laparoscopy following stab wounds to the flank or abdomen. Three of these patients were converted to laparotomy. One patient had a colon injury identified during laparoscopic evaluation, and this was repaired primarily following conversion to formal laparotomy. The second patient had equivocal findings at DL and the third had definitive evidence of peritoneal penetration without an identifiable intraabdominal injury. Both of these last two patients underwent open laparotomy which confirmed the laparoscopic findings. One patient underwent a therapeutic laparoscopy with repair of a diaphragmatic defect. Overall, 14 patients underwent DL following a penetrating trauma to the abdomen or flank without an

274 Table 3. Comparison of preoperative and intraoperative findings in laparotomy and laparoscopy patients Laparotomy Number of patients (n) Gunshot/shotgun Stab wound Preoperative indications Positive wound exploration Positive DPL Positive CT scan Evisceration Suspicion Intraoperative findings: No peritoneal violation Peritoneal violation without intraabdominal organ injury Intraabdominal organ injury not requiring intervention Intraabdominal organ injury requiring intervention 19 8 11 7 0 1 2 9 8 5 6 0 Laparoscopya 18 4 14 13 0 0 1 4 13 4 0 1

a Includes conversion patients undergoing both laparoscopy and laparotomy

identifiable intraabdominal injury requiring therapeutic intervention or conversion to laparotomy between January 1, 1992, and September 30, 1994 (Table 3). There were 11 patients with stab wounds and three patients with tangential gunshot wounds. Nine of 11 patients with stab wounds had preoperative evidence of peritoneal penetration with either an omental evisceration or a positive wound exploration. None of these patients had intraabdominal injuries identified at DL. The three patients with gunshot wounds were without signs of peritonitis prior to DL and no intraabdominal injuries were discovered during laparoscopic evaluation. Four patients, described previously, underwent both laparoscopy and laparotomy and are included in the CONV group. Nineteen patients who underwent negative or nontherapeutic laparotomy (NL) were identified during the same time period (Table 3). There were 11 patients with stab wounds, seven patients with gunshot wounds, and one patient with a shotgun wound in the NL group. Nine of 11 patients with stab wounds had either an omental evisceration or positive wound exploration. Three of the gunshot wound patients had signs of peritoneal irritation preoperatively. None of the stab wound victims had evidence of peritonitis. The one patient with a shotgun wound underwent a laparotomy following a CT scan which revealed a single intraabdominal pellet. This pellet was located in the omentum and no therapeutic intervention was necessary. There were no intraoperative or postoperative complications that prolonged, hospitalization or increased costs. Also, there were no missed injuries in the three groups. There was no significant difference in age, operative times, or ISS between the DL and NL groups (Table 4). Mean ISS of CONV patients was significantly greater than DL patients, 5.75 1.97 vs 2.43 0.63 (p < 0.05). Mean operative time for CONV patients was also significantly greater than both DL and NL patients, 106.5 17.00 min vs 66.1 6.55 and 47.3 7.50 min, respectively (p < 0.05). The mean length of stay was significantly shorter in the DL group as compared to the NL and CONV groups, 1.43 0.20 days, vs 4.26 0.31 days and 5.00 0.82 days, respectively (p < 0.0001). Seventy-three percent (10/14) of

patients undergoing DL were discharged within 1 day of surgery. One patient required hospitalization for 3 days because of an associated chest injury. In comparison, 94.7% of NL patients (18/19) remained hospitalized for 3 days or more. Variable costs reflect the number and nature of procedures undergone by each patient and include the cost of supplies and labor specific to each patients care. The physicians fees for each patient were added to both the variable and total costs. The variable costs for the DL group were significantly lower than those incurred by patients in the NL and CONV groups, $2,917 175 vs $3,384 102 and $3,774 286, (p < 0.05). Variable costs were not significantly different between the NL and CONV groups. Total costs were also significantly lower in the DL group when compared to NL and CONV, $5,427 394 vs $7,026 251 and $7,855 750 (p < 0.005), but again were not statistically different between the NL and CONV groups. A detailed analysis comparing the average costs incurred by laparoscopy and laparotomy is provided in Table 5. The overall total costs for laparoscopy, including the costs incurred by conversion patients, were significantly less than the total costs for laparotomy patients, $5,664 394 vs $7,028.47 250 (p < 0.005). This resulted in an overall savings of $1,059.44 per laparoscopy performed (Table 6). When the costs were divided into preoperative, intraoperative, and postoperative expenditures, the greatest proportion of costs in both the DL and NL groups was intraoperative, both in the analysis of variable and of total costs. Postoperative expenditures in the NL group, however, were almost three times greater than those incurred by patients in the laparoscopy group, reflecting the increased cost of prolonged hospitalization. Discussion The goal of our review was to assess the overall economic impact of diagnostic laparoscopy in the management of stable patients with penetrating abdominal trauma. The total costs calculated in this study, including the physicians fees, best represent the cost as perceived by a third-party payer. Alternatives to DL in the evaluation and management of hemodynamically stable patients with penetrating abdominal trauma include observation with selective laparotomy or mandatory laparotomy. We had no morbidity related to the laparoscopic procedures and demonstrated a significantly decreased length of stay. Several patients were discharged home immediately following DL without further observation, and over 70% were discharged by the 1 postoperative day. This would support the use of DL in the emergency department with local anesethesia and intravenous sedation [1, 2, 13]. Patients could be safely evaluated and discharged home without admission to the hospital. Salvino et al. have shown the safety and efficacy of emergency department laparoscopic evaluation [13]. In the ongoing evaluation of the use of DL in trauma patients, most protocols require taking the patient to the operating room, although not necessarily using general anesthesia [2, 3, 5, 8, 11, 15, 16]. It has been our practice, and it is our recommendation, to perform DL in the operating room setting. The use of the operating room affords the surgeon greater flexibility. Pa-

275 Table 4. Comparison of patients having undergone negative or nontherapeutic laparoscopy and/or laparotomy between January 1, 1992, and September 30, 1994 Negative or nontherapeutic laparoscopy Patients (n) Gunshot/shotgun victims (n) Stab wound victims (n) Mean age (years) Injury Severity Score Length of stay (days) Operative time (minutes) Variable costs (dollars) Total costs (dollars)
a b

Negative or nontherapeutic laparotomy

Laparoscopy and laparotomy (conversion) 4 1 3 of mean) 38.5 6.65 5.75 1.97 5.0 0.82 106.5 17.00 3774 286 7855 750

14 3 11

19 8 11 (Calculation of means standard error 31.2 2.23 3.21 0.66 4.26 0.31 66.1 6.55d 3384 102 7026 251

30.5 2.41 2.43 0.63a 1.43 0.20b 47.3 7.50c 2919 175e 5427 394f

p < 0.05 vs conversion p < 0.0001 vs laparotomy and conversion c p < 0.01 vs conversion d p < 0.05 vs conversion e p < 0.05 vs laparotomy and conversion f p < 0.001 vs laparotomy and conversion

tients may be easily repositioned and rotated on a standard operating room table to facilitate the complete evaluation of the abdomen. The ready availability of both 0 and 30 laparoscopes is also valuable. Lastly, in those patients who will require conversion to laparotomy due to inadequate laparoscopic evaluation or because of the need for therapeutic intervention, time will not be wasted transporting them to a different area of the hospital. The surgeon will be able to proceed directly to a laparotomy or to proceed with therapeutic laparoscopy. In addition, these procedures should be carried out by experienced laparoscopic surgeons who have familiarity with advanced laparoscopic techniques. Some of the reported studies in emergency department laparoscopy have also documented the use of smallersize laparoscopes [1, 2]. With the currently available 5-mm or 4-mm laparoscopes, however, visualization has been limited by a lack of brightness and a smaller field of vision. With newer technology and improvement in optics, the use of these smaller scopes may become more commonplace. There is no dispute regarding the efficacy and safety of DL in selected trauma patients. Carey et al. reported a decrease in their nontherapeutic laparotomy rate from 11% to 8.5% since initiating DL rather than diagnostic peritoneal lavage, computed tomography, or local wound exploration in the evaluation of patients with penetrating abdominal trauma [3]. The ability to decrease or even eliminate negative or nontherapeutic laparotomy could impact the overall morbidity and mortality of the trauma population [6, 10, 12]. Investigators at the University of Miami found that their 12.4% negative laparotomy rate following mandatory laparotomy was associated with a 22% morbidity rate in these patients and a mean hospital stay of 5.1 days [17]. By using diagnostic laparoscopy in a similar group of patients, they were able to decrease the morbidity rate to 3% and the mean hospital stay to 1.4 days. Both of these outcomes should significantly decrease the cost of caring for these patients. In our series, there were no missed injuries or complications. In addition, with the use of laparoscopy, we were able to significantly reduce our negative or nonthera-

peutic laparotomy rate from 33.9% to 19.1%. Laparoscopy proved to be not only safe, but it helped avoid an unnecessary laparotomy and the increased costs incurred by an extended hospitalization. Once it has been determined that a given procedure is safe and effective, the costs incurred by this procedure are commonly the next issue to be carefully evaluated. This study has demonstrated how decreased length of stay following DL leads to decreased total hospital costs, and we suspect that these costs will further decrease as the variable costs associated with laparoscopy decrease. The use of disposable instrumentation increases a hospitals variable costs. By eliminating the use of many disposable laparoscopic instruments, hospitals can reduce the costs of treating patients without affecting the standards of patient care. The use of gasless laparoscopy and conventional instruments has also been shown to be safe and cost effective [14]. This would not only lower the variable costs incurred by each patient but also lower the relative proportion of intraoperative costs associated with diagnostic or therapeutic laparoscopy. Also, by avoiding an unnecessary laparotomy and the increased costs incurred by this procedure, laparoscopy can provide further cost savings. Diagnostic laparoscopy is a cost-effective procedure for the evaluation of hemodynamically stable patients. It is vital that patients undergo a complete laparoscopic evaluation of the entire abdomen, preferentially in an operating room setting, where a surgeons resources are greatest. These procedures should all be carried out, or supervised, by experienced laparoscopic surgeons. In addition, all patients should consent to exploratory laparotomy, and there should be no hesitation to convert to an open procedure if the patient becomes unstable or if complete laparoscopic evaluation is not possible. In conclusion, hospital costs and length of stay were significantly lower in our population of patients when comparing negative or nontherapeutic laparoscopy with laparotomy in the evaluation of penetrating abdominal trauma. The overall economic impact of laparoscopy resulted in a $1,059 savings per laparoscopy performed.

276 Table 5. Detailed analysis of average costs incurred by laparoscopy and laparotomy Laparoscopy Variable Daily room and care IV therapy Emergency services Blood bank Chemistry lab Histology lab Hematology lab Electrocardiology lab Pharmacy Recovery room Operating room Central supply Respiratory care Anesthesia Microbiology lab Radiology Surgeons fee Anesthesiologists fee Total $244.67 6.92 94.67 26.00 46.42 0.00 13.17 2.00 76.25 51.33 171.25 634.67 9.50 253.83 1.75 28.08 950.00 308.77 $2,919.00 Total $843.42 13.67 246.17 66.50 110.08 0.00 40.42 5.42 142.33 120.33 869.58 1,122.67 24.33 435.25 3.83 124.67 950.00 308.77 $5,427.00 Variable $843.53 15.13 112.00 21.60 57.07 1.00 18.27 2.13 174.07 51.47 150.53 394.33 36.07 255.40 6.07 13.20 900.00 333.47 $3,384.00 Laparotomy Total $2,713.73 27.13 296.67 57.33 143.53 7.53 54.07 4.93 323.80 118.80 780.93 713.60 92.80 435.73 16.07 59.93 900.00 333.47 $7,026.00

Table 6. Comparison of overall costs incurred by laparoscopy during the study period vs the costs of negative or nontherapeutic laparotomy Laparoscopya Variable costs (dollars) Total costs (dollars)
a

6. Laparotomy 7. 3,384.37 102 7,028.47 251 8. 9. 10.

3,028.43 174b 5,664.57 394c

overall DL costs (DL costs + [CONV costs-NL costs])/DL + CONV patients (n) b p 0.07 c p < 0.005

Acknowledgment. We wish to thank Leslie Brown and Jane Dostal for their excellent assistance in the preparation of this manuscript.

11. 12. 13. 14. 15. 16. 17.

References
1. Berci G, Dunkelman D, Michel SL, Sanders G, Wahlstrom E, Morgenstern L (1983) Emergency minilaparoscopy in abdominal trauma. Am J Surg 146: 261265 2. Berci G, Sackier JM, Paz-Parlow M (1991) Emergency laparoscopy. Am J Surg 161: 332335 3. Carey JE, Koo R, Miller R, Stein M (1995) Laparoscopy and thoracoscopy in evaluation of abdominal trauma. Am Surg 61: 9295 4. Fabian TC, Croce MA, Stewart RM, Pritchard FE, Minard G, Kudsk KA (1993) A prospective analysis of diagnostic laparoscopy in trauma. Ann Surg 217: 557565 5. Fernando HC, Alle KM, Chen J, Davis I, Klein SR (1994) Triage by

laparoscopy in patients with penetrating abdominal trauma. Br J Surg 81: 384385 Henderson VJ, Organ CH Jr, Smith RS (1993) Negative trauma celiotomy. Am Surg 59: 365370 Ivatury RR, Simon RJ, Stahl WM (1993) A critical evaluation of laparoscopy in penetrating abdominal trauma. J Trauma 34: 822828 Livingston DH, Tortella BJ, Blackwood J, Machiedo GW, Rush BF (1992) The role of laparoscopy in abdominal trauma. J Trauma 33: 471475 Marks Jm, Ramey R, Baringer DC, Aszodi A, Ponsly JL (1995) Laparoscopic repair of a diaphragmatic laceration. Surg Laparosc Endosc 5: 415418 Petersen SR, Sheldon GF (1979) Morbidity of a negative finding at laparotomy in abdominal trauma. Surg Gynecol Obstet 148: 2326 Ponsky JL, Marks JM (1995) Laparoscopic examination of the bowel in trauma patients. Gastrointest Endosc 143(2): 146148 Ryan M, Leighton T, Pianim N, Klein S, Bongard F (1993) Medical economic consequences of gang-related shootings. Am Surg 59: 831 833 Salvino CK, Esposito TJ, Marshall WJ, Dries DJ, Morris RC, Gamelli RL (1993) The role of diagnostic laparoscopy in the management of trauma patients: a preliminary assessment. J Trauma 34: 506515 Smith RS, Fry WR, Tsoi EKM, Henderson VJ, Hirvela ER, Koehler RH, Brams DM, Morabito DJ, Peskin GW (1993) Gasless laparoscopy and conventional instruments. Arch Surg 128: 11021107 Smith RS, Fry WR, Morabito DJ, Koehler RH, Organ CH Jr. (1995) Therapeutic laparoscopy in trauma. Am J Surg 170: 632636 Sosa JL, Sims D, Martin L, Zeppa R (1992) Laparoscopic evaluation of tangential abdominal gunshot wounds. Arch Surg 127: 109110 Sosa JL, Baker M, Peunte I, Sims D, Sleeman D, Ginzburg E, Martin L (1995) Negative laparotomy in abdominal gunshot wounds: potential impact of laparoscopy. J Trauma 38: 194197