- Email: [email protected]
Changing patterns of treatment for blunt splenic injuries: An 11-year experience in a rural state
Changing Patterns of Treatment for Blunt Splenic Injuries: An 11-Year Experience in a Rural State By Carmine Frumiento and Dennis W. Vane Burlington, Vermont
Purpose: The aim of this study was to perform a populationbased study evaluating the trend in management of pediatric blunt splenic injuries in a rural state and assess differences in the management of those injuries at a level I pediatric trauma center (PTC) and regional hospitals (RH) from 1985 through 1995. Methods: ICD-9-CM diagnosis and procedure codes for children (age less than 19) discharged from all hospitals in a rural state with splenic injuries from 1985 through 1995 were reviewed. Hospital charges, age, and nonoperative management (NOM) rates were calculated for PTC and RH and compared using 2 and linear regression. (P ⬍ .05 is statistically significant.) Patients were divided into 2 groups; G1, 1985 through 1989 (127 children); G2, 1990 through 1995 (140 children). Results: The overall NOM rate increased from 21% (G1) to 64.2% (G2), P ⬍ .001 A total of 114 patients were treated at PTC and 153 patients received care at RH. PTC had a NOM rate of 54.3% versus 35.9% at RH (P ⫽ .003). There was no statistical difference in ages or ISS within the groups or
T
HE SPLEEN is the most commonly injured intraabdominal solid organ after blunt abdominal trauma. Up until the late 1980s the standard of care for blunt splenic injuries was splenectomy. After the description of overwhelming postsplenectomy sepsis (OPSI) in children by King and Shumacker in 1952, investigators searched for alternatives to splenectomy including both splenorrhaphy and nonoperative management (NOM) in hemodynamically stable children.1,2 The trend toward nonoperative management of blunt splenic injuries in stable children first began in the late 1960s with a report from Upadhyaya and Simpson.3 Since that time the safety and efficacy of this modality in the treatment hemodynamically stable children with blunt splenic injuries has been validated by numerous investigators in the pediatric literature and has become the standard of care.4-7 Following the lead from the pediatric literature, the successful nonoperative management of adult blunt splenic injury increased as well with multiple studies supporting its efficacy in the stable adult population.8-12 Despite the acceptance of nonoperative management for the treatment of blunt splenic injuries, differences still exist between the aggressiveness of adult surgeons treating trauma in children and pediatric trauma surgeons in instituting nonoperative management. A report from our Journal of Pediatric Surgery, Vol 35, No 6 (June), 2000: pp 985-989
between PTC and RH. NOM in RH rose from 7.7% in G1 to 56.9% in G2 (P ⬍ .000), and from 35.5% in G1 to 76.9% in G2 (P ⬍ .001) for PTC. Hospital charges were lower for patients receiving NOM versus those with surgical treatment of their injury, $8,094 versus $10,862 (P ⫽ .018). However, a higher percentage of children were treated at RH than PTC in G2 versus G1 (68.2% v 51.2%, P ⫽ .0541).
Conclusions: Over the 10-year period studied, the NOM rate for splenic injuries significantly decreased. This trend was seen at both the PTC and RH, but the PTC maintained a higher rate of NOM. Unfortunately, more children were treated at RH in G2. Educational programs increased NOM in RH but not to a level equal to PTC. These programs had the negative effect of allowing more children to be treated at RH, actually increasing the splenic operation rate for this population. J Pediatr Surg 35:985-989. Copyright 娀 2000 by W.B. Saunders Company.
INDEX WORDS: Splenic injury, nonoperative management.
institution by Keller and Vane13 in 1995 found that pediatric trauma surgeons treated significantly more of their patients with blunt splenic injuries nonoperatively then did adult trauma surgeons treating children with similar injuries. The purpose of this study was to evaluate the trend in management of pediatric blunt splenic injuries in a rural state and assess differences in the management of these injuries at a level I pediatric trauma center (PTC) and regional hospitals (RH) throughout the state from 1985 to 1995. In addition, we were interested in the effect that educational programs about NOM, instituted in 1990 within our state, had on increasing the rates of nonoperative management at both PTC and RH over this period. Our hypothesis was that educational programs increase
From the University of Vermont College of Medicine, the Department of Surgery, Fletcher Allen Health Care, Burlington, VT. Presented at the 1999 Annual Meeting of the Section on Surgery of the American Academy of Pediatrics, Washington, DC, October 8-10, 1999. Address reprint requests to Dennis W. Vane, MD, Department of Surgery, University of Vermont, Burlington, VT 05401. Copyright 娀 2000 by W.B. Saunders Company 0022-3468/00/3506-0034$03.00/0 doi:10.1053/js.2000.6948 985
986
FRUMIENTO AND VANE
Table 3. Hospital Charges: Operative Versus Nonoperative Treatment of Splenic Injuries
Table 1. Nonoperative Rates: Group 1 Versus Group 2
NOM rate (statewide) RH-NOM rate PTC-NOM rate
Group 1 (n ⫽ 127) (1985-1989)
Group 2 (n ⫽ 140) (1990-1995)
P Value
27/127 (21%) 5/65 (7.7%) 22/62 (35.5%)
90/140 (64.2%) 50/88 (56.9%) 40/52 (76.9%)
⬍.001 ⬍.001 ⬍.001
the percentage of pediatric splenic injuries treated with nonoperative management at both PTC and RH over the past 11 years, but PTC continue to have a higher rate of NOM of blunt splenic injuries. MATERIALS AND METHODS ICD-9-cm diagnosis and procedure codes for all children less than 19 years of age discharged from all hospitals in a rural state with splenic injuries from 1985 through 1995 were reviewed. Children were divided into 2 groups: group 1 treated in the time period 1985 through 1989, and group 2 treated in the time period 1990 through 1995. Hospital charges, average age, injury severity score (ISS), and nonoperative management rates were calculated for PTC and RH and compared using 2 and linear regression. A P value less than .05 was considered statistically significant. An educational program dealing with NOM of blunt abdominal injuries was instituted within the state in 1990. It encompassed multiple site visits and lectures by a pediatric trauma surgeon to appropriate physicians and staff at each of the hospitals throughout the state.
RESULTS
There were 267 children included in this study. One hundred twenty-seven were in group 1 and 140 were in group 2. The average age for children treated at the PTC was 13.1 years, whereas for those treated at RH was 13.6 years (P value not significant). The overall NOM rate increased significantly throughout the state from 21% in group 1 to 64% in group 2 (P ⬍ .001). A total of 114 children were treated at the PTC (62 in group 1 and 52 in group 2), whereas 153 children received their treatment at RH (65 in group 1 and 88 in group 2) over this 11-year period. The PTC had an overall NOM rate of 54.3% versus 35.9% for RH (P ⫽ .003). There also were significant differences in NOM rates when comparing PTC versus RH in group 1 only and in group 2 only. The NOM rate for PTC in group 1 was 35.5% versus 7.7% for RH (P ⬍ .001). Although the NOM rate for PTC in group 2 was 77% versus 57% for the RH (P ⫽ .016; Tables 1, 2, and 3). RH increased their rate of NOM from 7.7% in group 1 Table 2. Nonoperative Rates: PTC Versus RH PTC
No. of patients NOM rate (overall) NOM rate (group 1) NOM rate (group 2) Location of treatment Group 1 Location of treatment Group 2
RH
114 62/114 (54.3%) 22/62 (35.5%) 40/52 (76.9%)
153 55/153 (35.9%) 5/65 (7.7%) 50/88 (56.9%)
62/127 (48.8%)
65/127 (51.2%)
52/140 (37.2%)
88/140 (62.8%)
P Value
.003 ⬍.001 .016
.054
Average hospital charges
Operative Treatment
NOM Treatment
P Value
$8,094
$10,962
.018
to 56.9% in group 2 (P ⬍ .001). At the PTC, the rate of NOM also increased significantly for treatment of splenic injuries from 35.5% in group 1 to 76.9% in group 2 (P ⬍ .001). The regression plots in Figs 1 and 2 show this decrease in surgical management of blunt splenic injuries at both PTC and RH over this period. Figure 3 compares NOM rates at PTC and RH by year. The mean ISS for group 1 was 9.9 for RH and 15.1 for PTC, and the mean ISS for group 2 was 13.9 for RH and 16.3 for PTC. Average hospital charges were significantly lower for children receiving NOM versus surgical management of their splenic injuries. ($8,094 v $10,862, respectively; P ⫽ .018). The percentage of children receiving care at RH and at PTC was nearly identical in group 1. A total of 65 (51.2%) children were treated for splenic injuries at RH, whereas 62 (48.8%) received treatment at PTC. However, the percentage of children receiving treatment at RH versus PTC in group 2 increased to a ratio of nearly 2:1. Eighty-eight of 140 (62.8%) children in group 2 were treated at RH, whereas 52 (37.2%) were treated at PTC (P ⫽ .0541). There were 5 deaths in this study, 4 at the PTC, and 1 at RH. All children that died sustained massive head injuries in addition to their splenic injury. DISCUSSION
The spleen is the most commonly injured intrabdominal solid organ. After the first successful report of splenectomy for blunt splenic trauma in the late 19th century, splenectomy quickly became the treatment of choice in all splenic injuries up to the late 20th century. However, splenectomy began to be questioned as the best treatment in blunt splenic injuries after the first report of overwhelming postsplenectomy sepsis by King and Shumacker in 1952.1 This was followed by the first report of NOM after blunt splenic injuries in children in the late 1960s. Since that time multiple findings have confirmed the efficacy and safety of NOM in such injuries in children with success rates of NOM upwards of 90%, thus making NOM in hemodynamically stable children with blunt splenic injuries the standard of care.3-7 More recently, adult trauma surgeons have come to accept NOM of hemodynamically stable patients with similar injuries. However, the rate at which adult trauma surgeons treat blunt splenic injuries in children nonoperatively versus that of pediatric surgeons was found to be significantly different in a prior study by Keller and Vane13 from our institution in 1995. They found that 79%
BLUNT SPLENIC INJURIES: CHANGING PATTERNS OF CARE
Fig 1. Percent operative management versus time at PTC. %OPPTC, percent of patients treated operatively at pediatric trauma center.
of children with blunt splenic injuries cared for by pediatric surgeons were treated nonoperatively versus only 48% of children cared for by adult trauma surgeons.13 Preliminary findings in accordance with the final results led to an attempt to increase the rate of NOM for stable children with bunt splenic injury throughout our rural state. To this end, in 1990, pediatric surgeons from the state’s only level 1 pediatric trauma center began having site visits and providing information and lectures about NOM in stable children to general surgeons and appropriate staff at regional hospitals throughout the state. This study was undertaken to evaluate any trend changes in the management of stable children with blunt splenic injuries within our state. Overall, more children were treated at regional hospitals for their splenic injuries than at the PTC. The NOM rate for the entire state in group 2 of 64.2% is comparable to the finding in a study by Hunt et al14 in 1995 of a 62.9% nonoperative treatment rate during a similar time period. Group 1 children treated at RH were almost exclusively managed operatively versus only about a one third of
Fig 2. Percent operative management versus time at RH. %OP-RH, Percent of patients treated operatively at regional hospitals.
987
Fig 3. Nonoperative management rates for PTC versus RH. %NOMPTC, percent of patients with blunt splenic injuries treated nonoperatively at the pediatric trauma center. %NOM-RH, percent of patients with blunt splenic injuries treated nonoperatively at regional hospitals.
patients managed nonoperatively at PTC for the same period. These rates improved significantly at both the PTC and RH in group 2, reaching a high of 100% of 10 patients with blunt splenic injuries being managed nonoperatively at the PTC in 1995 as shown on Fig 3. Figure 3 also shows a 0% nonoperative rate for PTC in 1991. There was 1 child treated at the PTC that year with a blunt spleen injury, and he received a splenectomy. Although finding only 1 patient was suspicious to us at first, this was confirmed by manual review of CT scan logs that found no other children with splenic injuries diagnosed during that year at the PTC. The rate of NOM was significantly higher at the PTC than at the RH for both group 1 and group 2 as was be expected. However, although the PTC doubled its rate of NOM, RH had a remarkable 8-fold increase from 7.7% to 56.9% in their rate of NOM for these injuries. Although proving that the educational programs instituted in 1990 were solely responsible for this increase would have required a control group of similar hospitals, which was not feasible, undoubtedly these programs did contribute to bring about this change. In conjunction with individual general surgeons at RH keeping current on their own about treatment of splenic injuries in children, the pediatric surgeons provided an additional resource of easily available information and experience. The mean ISS for both the RH and the PTC increased during the 2 measured periods. The reason for this is somewhat unclear but may be the result of the RH keeping children with higher ISS at their facilities. Subsequently, some of the lesser injured children previously transferred to the PTC remain at the RH. As a result, the ISS in both groups increases. In today’s health care environment, financial issues are receiving greater consideration than ever before in the type of care a patient receives. The goals are to provide
988
FRUMIENTO AND VANE
the best care in the most efficient way possible. Previous studies have shown that NOM of stable children after blunt splenic injury actually can be more cost effective than surgical management and that the length of hospital stay was not significantly greater.13 This current study confirms this cost benefit at least indirectly by showing a significant difference in hospital charges. The most unexpected finding from this study was that a greater percentage of children in group 2 are receiving their care at RH rather than PTC. Between 1985 and 1989, 48.8% of the children with splenic injuries received their care at the PTC versus 51.2% who received care at the RH. From 1990 through 1995, the percentage of children receiving care at the RH increased to 62.8%, whereas only 37.2% of the children are now being treated at the PTC. Although this is not statistically significant, it does approach statistical significance with a P value of .054. A larger population of patients would be needed to show a statistical significant difference. Interestingly, a hypothetical model using the group 2 NOM rates for PTC and RH while using the group I distribution of treatment sites would result in 3 fewer children having surgery. This clearly is a small number, but if this trend increases, the number may become significant. Also, if this shows a national trend, the number of children receiving splenectomy may become inappropriately high.
As we all know, nonoperative management does not mean no treatment, on the contrary, successful NOM demands continuous monitoring in an intensive care unit setting for at least the first 24 to 48 hours as well as supervision of patients with particular attention to changes in physical examination findings. In addition, there should be rapid availability of surgical resources. It would seem that a tertiary care center is better set up for the demands of such care, and we actually had expected the to see an increase in the number of children transferred to the PTC for their splenic injury rather than the drop in percentage that we found. This study found that the rate of nonoperative management of blunt splenic injuries in children increased significantly from the group 1 patients to the group 2 patients at both the PTC and RH. However, PTC still treat significantly more of their children with splenic injuries nonoperatively than do RH within the state. Educational programs instituted by pediatric surgeons at the PTC likely played a role in the 8-fold increase in the rate of NOM at RH. Surprisingly, a greater percentage of children within the state are receiving care for splenic injuries at RH rather than at PTC. This is despite what would appear to be a more labor intensive treatment option for which the PTC is likely better equipped to carry out.
REFERENCES 1. King H, Shumacker HB Jr: Splenic studies I: Susceptibility to infection after splenectomy performed in infancy. Ann Surg 136:239242, 1952 2. O’Connor GS, Geelhoed GW: Splenic trauma and salvage. Am Surg 52:456-462, 1986 3. Upadhyaya P, Simpson JS: Splenic trauma in children. Surg Gynecol Obstet 126:781-790, 1969 4. Ein SH, Shandling B, Simpson JS, et al: Non-operative management of traumatized spleen in children: How and why. J Pediatr Surg 13:117-119, 1978 5. Morse MA, Garcia VF: Selective non-operative management of pediatric blunt splenic trauma: Risk for missed associated injuries. J Pediatr Surg 29:23-27, 1994 6. Heller JA Jr, Papa P, Drugas G, et al: Non-operative management of solid organ injuries in children—Is it safe? Ann Surg 219:625-631, 1994 7. Sjovall A, Hirsh K: Blunt abdominal trauma in Children: Risks of non-operative treatment. J Pediatr Surg 32:1169-1174, 1997 8. Cogbill TH, Moore EE, Jurkovick GJ, et al: Nonoperative
management of blunt splenic trauma: A multi-center experience. J Trauma 29:1312-1317, 1989 9. Powell M, Courcoulas A, Gardner M, et al: Management of blunt splenic trauma: Significant differences between adults and children. Surgery 122:654-660, 1997 10. Bianchi JD, Collin GR: Management of splenic trauma at a rural, level I trauma center. Am Surg 63:490-495, 1997 11. Archer LP, Rogers FB, Shackford SR: Selective nonoperative management of liver and spleen injuries in neurologically impaired adult patients. Arch Surg 131:309-315, 1996 12. Villalba MR, Howells GA, Lucas RJ, et al: Non-operative management of the adult ruptured spleen. Arch Surg 125:836-840, 1990 13. Keller MS, Vane DW: Management of pediatric blunt splenic injury: The comparison of pediatric and adult trauma surgeons. J Pediatr Surg 30:221-225, 1995 14. Hunt JP, Lentz CW, Cairns BA, et al: Management and outcome of splenic injury: The results of a five year state wide population-based study. Am Surg 62:911-917, 1996
Discussion D. Mooney (Boston, MA): The increase in nonoperative management appeared to occur to regional hospitals in 1987 before your educational program was instituted, and I was wondering if you could elaborate on that? Secondly, at your institution, my understanding is that
the adult surgeons manage either the 16 or 17 year olds. I wonder why you included patients treated at your institution by the adult surgeons in your data set instead of just purely the patients treated by pediatric surgeons. Could you comment on the number of patients admit-
BLUNT SPLENIC INJURIES: CHANGING PATTERNS OF CARE
ted to your facility with this injury who are transferred versus admitted directly from the scene? It may be that the more stable patients were transferred to a facility and the other patients were treated at the referring hospital. Data studies from the State of Vermont, when they talk about transfer to a pediatric trauma center, need to consider the increase from 1993 through 1999 and the number of patients from Vermont transferred to Dartmouth in your manipulations. Finally, you imply in your data that your institution is a Level 1 pediatric trauma center and was through this study period. Could you just comment on how much of the study time your institution was actually a College of Surgeons Level 1 pediatric trauma center? C. Frumiento (response): I believe the first question dealt with when we instituted our program in 1990 and we start to see an increase in 1987. Actually, our graph does not show a very significant increase before 1990 in the regional hospitals. Our educational programs I think played a role in this. But we would need a control group in which no educational programs were done to truly prove that they were the sole reason for this increase. And we do not make that claim that the educational programs were the sole reason for the increase. We believe that physicians working around the state do try to stay current on the literature, and we just provide an additional resource to give them information as to how to manage these children. Your second question had to do with adults. It sort of gets mixed, and it gets a little blended in our institution. We like to say that 17 years and younger go to the pediatric surgery service. However, there are many instances in which we have had 18 and 19 year olds in our service. It is a good point. And we could look back and probably remove a handful of patients from this study group if we look specifically at 17 and under. As for transferring patients who are more stable, that is another very good point. But you also will notice that the ISS at the pediatric trauma centers were actually higher. That does not necessarily mean they were less stable, just
989
potentially more injured, perhaps had a higher score of injury. So it sort of cuts both ways. Yes, they might have kept and operated on some patients that were more unstable, but they also might have transferred some of those patients to a tertiary care center. As far as the level 1 status of our institution, I believe it gained level 1 status in the early 1990s. To that end, it was still a tertiary referral center for injured children with pediatric surgeons on staff throughout the study period. As far as Dartmouth, you are absolutely right, some children were referred to Dartmouth over the study period. Our database included all Vermont citizens, if you will, so we actually do have the data for patients that were transferred to Dartmouth. And in the last 5 years there were a handful, about 6 to 8 patients that were transferred to Dartmouth for care. That would change that P value, would raise the P value somewhat on that comparison looking at treatment at regional hospitals versus pediatric trauma center, increasing the number of patients that were referred to a trauma center. However, it would still leave the trend intact. With more patients, I think it would become statistically significant. R. Pearl (Peoria, IL): Your 70% plus nonoperative rate at Children’s Hospital actually is a very low. Most published series in the mid 1980s were about 85%, in the 1990s it is over 90%, with splenic salvage rates with splenorrhaphy of 95% to 97%. I wonder, are your rates as low as they are because general surgeons are caring for them at your center, you are including 19-year-olds, or do you have a population with higher ISS and therefore sometimes you do splenectomies because, for example, they had a perforated duodenum? C. Frumiento (response): These are very good questions. Our rates in the study period 1992 through 1995 were actually consistently above 85%, in the last several years in the study were above 90% looking at our graph. Before that, the rates were somewhat on the low side. But again, you mentioned it yourself, the ISS were a little higher than some of the ISS quoted in these other studies.
Purpose: The aim of this study was to perform a populationbased study evaluating the trend in management of pediatric blunt splenic injuries in a rural state and assess differences in the management of those injuries at a level I pediatric trauma center (PTC) and regional hospitals (RH) from 1985 through 1995. Methods: ICD-9-CM diagnosis and procedure codes for children (age less than 19) discharged from all hospitals in a rural state with splenic injuries from 1985 through 1995 were reviewed. Hospital charges, age, and nonoperative management (NOM) rates were calculated for PTC and RH and compared using 2 and linear regression. (P ⬍ .05 is statistically significant.) Patients were divided into 2 groups; G1, 1985 through 1989 (127 children); G2, 1990 through 1995 (140 children). Results: The overall NOM rate increased from 21% (G1) to 64.2% (G2), P ⬍ .001 A total of 114 patients were treated at PTC and 153 patients received care at RH. PTC had a NOM rate of 54.3% versus 35.9% at RH (P ⫽ .003). There was no statistical difference in ages or ISS within the groups or
T
HE SPLEEN is the most commonly injured intraabdominal solid organ after blunt abdominal trauma. Up until the late 1980s the standard of care for blunt splenic injuries was splenectomy. After the description of overwhelming postsplenectomy sepsis (OPSI) in children by King and Shumacker in 1952, investigators searched for alternatives to splenectomy including both splenorrhaphy and nonoperative management (NOM) in hemodynamically stable children.1,2 The trend toward nonoperative management of blunt splenic injuries in stable children first began in the late 1960s with a report from Upadhyaya and Simpson.3 Since that time the safety and efficacy of this modality in the treatment hemodynamically stable children with blunt splenic injuries has been validated by numerous investigators in the pediatric literature and has become the standard of care.4-7 Following the lead from the pediatric literature, the successful nonoperative management of adult blunt splenic injury increased as well with multiple studies supporting its efficacy in the stable adult population.8-12 Despite the acceptance of nonoperative management for the treatment of blunt splenic injuries, differences still exist between the aggressiveness of adult surgeons treating trauma in children and pediatric trauma surgeons in instituting nonoperative management. A report from our Journal of Pediatric Surgery, Vol 35, No 6 (June), 2000: pp 985-989
between PTC and RH. NOM in RH rose from 7.7% in G1 to 56.9% in G2 (P ⬍ .000), and from 35.5% in G1 to 76.9% in G2 (P ⬍ .001) for PTC. Hospital charges were lower for patients receiving NOM versus those with surgical treatment of their injury, $8,094 versus $10,862 (P ⫽ .018). However, a higher percentage of children were treated at RH than PTC in G2 versus G1 (68.2% v 51.2%, P ⫽ .0541).
Conclusions: Over the 10-year period studied, the NOM rate for splenic injuries significantly decreased. This trend was seen at both the PTC and RH, but the PTC maintained a higher rate of NOM. Unfortunately, more children were treated at RH in G2. Educational programs increased NOM in RH but not to a level equal to PTC. These programs had the negative effect of allowing more children to be treated at RH, actually increasing the splenic operation rate for this population. J Pediatr Surg 35:985-989. Copyright 娀 2000 by W.B. Saunders Company.
INDEX WORDS: Splenic injury, nonoperative management.
institution by Keller and Vane13 in 1995 found that pediatric trauma surgeons treated significantly more of their patients with blunt splenic injuries nonoperatively then did adult trauma surgeons treating children with similar injuries. The purpose of this study was to evaluate the trend in management of pediatric blunt splenic injuries in a rural state and assess differences in the management of these injuries at a level I pediatric trauma center (PTC) and regional hospitals (RH) throughout the state from 1985 to 1995. In addition, we were interested in the effect that educational programs about NOM, instituted in 1990 within our state, had on increasing the rates of nonoperative management at both PTC and RH over this period. Our hypothesis was that educational programs increase
From the University of Vermont College of Medicine, the Department of Surgery, Fletcher Allen Health Care, Burlington, VT. Presented at the 1999 Annual Meeting of the Section on Surgery of the American Academy of Pediatrics, Washington, DC, October 8-10, 1999. Address reprint requests to Dennis W. Vane, MD, Department of Surgery, University of Vermont, Burlington, VT 05401. Copyright 娀 2000 by W.B. Saunders Company 0022-3468/00/3506-0034$03.00/0 doi:10.1053/js.2000.6948 985
986
FRUMIENTO AND VANE
Table 3. Hospital Charges: Operative Versus Nonoperative Treatment of Splenic Injuries
Table 1. Nonoperative Rates: Group 1 Versus Group 2
NOM rate (statewide) RH-NOM rate PTC-NOM rate
Group 1 (n ⫽ 127) (1985-1989)
Group 2 (n ⫽ 140) (1990-1995)
P Value
27/127 (21%) 5/65 (7.7%) 22/62 (35.5%)
90/140 (64.2%) 50/88 (56.9%) 40/52 (76.9%)
⬍.001 ⬍.001 ⬍.001
the percentage of pediatric splenic injuries treated with nonoperative management at both PTC and RH over the past 11 years, but PTC continue to have a higher rate of NOM of blunt splenic injuries. MATERIALS AND METHODS ICD-9-cm diagnosis and procedure codes for all children less than 19 years of age discharged from all hospitals in a rural state with splenic injuries from 1985 through 1995 were reviewed. Children were divided into 2 groups: group 1 treated in the time period 1985 through 1989, and group 2 treated in the time period 1990 through 1995. Hospital charges, average age, injury severity score (ISS), and nonoperative management rates were calculated for PTC and RH and compared using 2 and linear regression. A P value less than .05 was considered statistically significant. An educational program dealing with NOM of blunt abdominal injuries was instituted within the state in 1990. It encompassed multiple site visits and lectures by a pediatric trauma surgeon to appropriate physicians and staff at each of the hospitals throughout the state.
RESULTS
There were 267 children included in this study. One hundred twenty-seven were in group 1 and 140 were in group 2. The average age for children treated at the PTC was 13.1 years, whereas for those treated at RH was 13.6 years (P value not significant). The overall NOM rate increased significantly throughout the state from 21% in group 1 to 64% in group 2 (P ⬍ .001). A total of 114 children were treated at the PTC (62 in group 1 and 52 in group 2), whereas 153 children received their treatment at RH (65 in group 1 and 88 in group 2) over this 11-year period. The PTC had an overall NOM rate of 54.3% versus 35.9% for RH (P ⫽ .003). There also were significant differences in NOM rates when comparing PTC versus RH in group 1 only and in group 2 only. The NOM rate for PTC in group 1 was 35.5% versus 7.7% for RH (P ⬍ .001). Although the NOM rate for PTC in group 2 was 77% versus 57% for the RH (P ⫽ .016; Tables 1, 2, and 3). RH increased their rate of NOM from 7.7% in group 1 Table 2. Nonoperative Rates: PTC Versus RH PTC
No. of patients NOM rate (overall) NOM rate (group 1) NOM rate (group 2) Location of treatment Group 1 Location of treatment Group 2
RH
114 62/114 (54.3%) 22/62 (35.5%) 40/52 (76.9%)
153 55/153 (35.9%) 5/65 (7.7%) 50/88 (56.9%)
62/127 (48.8%)
65/127 (51.2%)
52/140 (37.2%)
88/140 (62.8%)
P Value
.003 ⬍.001 .016
.054
Average hospital charges
Operative Treatment
NOM Treatment
P Value
$8,094
$10,962
.018
to 56.9% in group 2 (P ⬍ .001). At the PTC, the rate of NOM also increased significantly for treatment of splenic injuries from 35.5% in group 1 to 76.9% in group 2 (P ⬍ .001). The regression plots in Figs 1 and 2 show this decrease in surgical management of blunt splenic injuries at both PTC and RH over this period. Figure 3 compares NOM rates at PTC and RH by year. The mean ISS for group 1 was 9.9 for RH and 15.1 for PTC, and the mean ISS for group 2 was 13.9 for RH and 16.3 for PTC. Average hospital charges were significantly lower for children receiving NOM versus surgical management of their splenic injuries. ($8,094 v $10,862, respectively; P ⫽ .018). The percentage of children receiving care at RH and at PTC was nearly identical in group 1. A total of 65 (51.2%) children were treated for splenic injuries at RH, whereas 62 (48.8%) received treatment at PTC. However, the percentage of children receiving treatment at RH versus PTC in group 2 increased to a ratio of nearly 2:1. Eighty-eight of 140 (62.8%) children in group 2 were treated at RH, whereas 52 (37.2%) were treated at PTC (P ⫽ .0541). There were 5 deaths in this study, 4 at the PTC, and 1 at RH. All children that died sustained massive head injuries in addition to their splenic injury. DISCUSSION
The spleen is the most commonly injured intrabdominal solid organ. After the first successful report of splenectomy for blunt splenic trauma in the late 19th century, splenectomy quickly became the treatment of choice in all splenic injuries up to the late 20th century. However, splenectomy began to be questioned as the best treatment in blunt splenic injuries after the first report of overwhelming postsplenectomy sepsis by King and Shumacker in 1952.1 This was followed by the first report of NOM after blunt splenic injuries in children in the late 1960s. Since that time multiple findings have confirmed the efficacy and safety of NOM in such injuries in children with success rates of NOM upwards of 90%, thus making NOM in hemodynamically stable children with blunt splenic injuries the standard of care.3-7 More recently, adult trauma surgeons have come to accept NOM of hemodynamically stable patients with similar injuries. However, the rate at which adult trauma surgeons treat blunt splenic injuries in children nonoperatively versus that of pediatric surgeons was found to be significantly different in a prior study by Keller and Vane13 from our institution in 1995. They found that 79%
BLUNT SPLENIC INJURIES: CHANGING PATTERNS OF CARE
Fig 1. Percent operative management versus time at PTC. %OPPTC, percent of patients treated operatively at pediatric trauma center.
of children with blunt splenic injuries cared for by pediatric surgeons were treated nonoperatively versus only 48% of children cared for by adult trauma surgeons.13 Preliminary findings in accordance with the final results led to an attempt to increase the rate of NOM for stable children with bunt splenic injury throughout our rural state. To this end, in 1990, pediatric surgeons from the state’s only level 1 pediatric trauma center began having site visits and providing information and lectures about NOM in stable children to general surgeons and appropriate staff at regional hospitals throughout the state. This study was undertaken to evaluate any trend changes in the management of stable children with blunt splenic injuries within our state. Overall, more children were treated at regional hospitals for their splenic injuries than at the PTC. The NOM rate for the entire state in group 2 of 64.2% is comparable to the finding in a study by Hunt et al14 in 1995 of a 62.9% nonoperative treatment rate during a similar time period. Group 1 children treated at RH were almost exclusively managed operatively versus only about a one third of
Fig 2. Percent operative management versus time at RH. %OP-RH, Percent of patients treated operatively at regional hospitals.
987
Fig 3. Nonoperative management rates for PTC versus RH. %NOMPTC, percent of patients with blunt splenic injuries treated nonoperatively at the pediatric trauma center. %NOM-RH, percent of patients with blunt splenic injuries treated nonoperatively at regional hospitals.
patients managed nonoperatively at PTC for the same period. These rates improved significantly at both the PTC and RH in group 2, reaching a high of 100% of 10 patients with blunt splenic injuries being managed nonoperatively at the PTC in 1995 as shown on Fig 3. Figure 3 also shows a 0% nonoperative rate for PTC in 1991. There was 1 child treated at the PTC that year with a blunt spleen injury, and he received a splenectomy. Although finding only 1 patient was suspicious to us at first, this was confirmed by manual review of CT scan logs that found no other children with splenic injuries diagnosed during that year at the PTC. The rate of NOM was significantly higher at the PTC than at the RH for both group 1 and group 2 as was be expected. However, although the PTC doubled its rate of NOM, RH had a remarkable 8-fold increase from 7.7% to 56.9% in their rate of NOM for these injuries. Although proving that the educational programs instituted in 1990 were solely responsible for this increase would have required a control group of similar hospitals, which was not feasible, undoubtedly these programs did contribute to bring about this change. In conjunction with individual general surgeons at RH keeping current on their own about treatment of splenic injuries in children, the pediatric surgeons provided an additional resource of easily available information and experience. The mean ISS for both the RH and the PTC increased during the 2 measured periods. The reason for this is somewhat unclear but may be the result of the RH keeping children with higher ISS at their facilities. Subsequently, some of the lesser injured children previously transferred to the PTC remain at the RH. As a result, the ISS in both groups increases. In today’s health care environment, financial issues are receiving greater consideration than ever before in the type of care a patient receives. The goals are to provide
988
FRUMIENTO AND VANE
the best care in the most efficient way possible. Previous studies have shown that NOM of stable children after blunt splenic injury actually can be more cost effective than surgical management and that the length of hospital stay was not significantly greater.13 This current study confirms this cost benefit at least indirectly by showing a significant difference in hospital charges. The most unexpected finding from this study was that a greater percentage of children in group 2 are receiving their care at RH rather than PTC. Between 1985 and 1989, 48.8% of the children with splenic injuries received their care at the PTC versus 51.2% who received care at the RH. From 1990 through 1995, the percentage of children receiving care at the RH increased to 62.8%, whereas only 37.2% of the children are now being treated at the PTC. Although this is not statistically significant, it does approach statistical significance with a P value of .054. A larger population of patients would be needed to show a statistical significant difference. Interestingly, a hypothetical model using the group 2 NOM rates for PTC and RH while using the group I distribution of treatment sites would result in 3 fewer children having surgery. This clearly is a small number, but if this trend increases, the number may become significant. Also, if this shows a national trend, the number of children receiving splenectomy may become inappropriately high.
As we all know, nonoperative management does not mean no treatment, on the contrary, successful NOM demands continuous monitoring in an intensive care unit setting for at least the first 24 to 48 hours as well as supervision of patients with particular attention to changes in physical examination findings. In addition, there should be rapid availability of surgical resources. It would seem that a tertiary care center is better set up for the demands of such care, and we actually had expected the to see an increase in the number of children transferred to the PTC for their splenic injury rather than the drop in percentage that we found. This study found that the rate of nonoperative management of blunt splenic injuries in children increased significantly from the group 1 patients to the group 2 patients at both the PTC and RH. However, PTC still treat significantly more of their children with splenic injuries nonoperatively than do RH within the state. Educational programs instituted by pediatric surgeons at the PTC likely played a role in the 8-fold increase in the rate of NOM at RH. Surprisingly, a greater percentage of children within the state are receiving care for splenic injuries at RH rather than at PTC. This is despite what would appear to be a more labor intensive treatment option for which the PTC is likely better equipped to carry out.
REFERENCES 1. King H, Shumacker HB Jr: Splenic studies I: Susceptibility to infection after splenectomy performed in infancy. Ann Surg 136:239242, 1952 2. O’Connor GS, Geelhoed GW: Splenic trauma and salvage. Am Surg 52:456-462, 1986 3. Upadhyaya P, Simpson JS: Splenic trauma in children. Surg Gynecol Obstet 126:781-790, 1969 4. Ein SH, Shandling B, Simpson JS, et al: Non-operative management of traumatized spleen in children: How and why. J Pediatr Surg 13:117-119, 1978 5. Morse MA, Garcia VF: Selective non-operative management of pediatric blunt splenic trauma: Risk for missed associated injuries. J Pediatr Surg 29:23-27, 1994 6. Heller JA Jr, Papa P, Drugas G, et al: Non-operative management of solid organ injuries in children—Is it safe? Ann Surg 219:625-631, 1994 7. Sjovall A, Hirsh K: Blunt abdominal trauma in Children: Risks of non-operative treatment. J Pediatr Surg 32:1169-1174, 1997 8. Cogbill TH, Moore EE, Jurkovick GJ, et al: Nonoperative
management of blunt splenic trauma: A multi-center experience. J Trauma 29:1312-1317, 1989 9. Powell M, Courcoulas A, Gardner M, et al: Management of blunt splenic trauma: Significant differences between adults and children. Surgery 122:654-660, 1997 10. Bianchi JD, Collin GR: Management of splenic trauma at a rural, level I trauma center. Am Surg 63:490-495, 1997 11. Archer LP, Rogers FB, Shackford SR: Selective nonoperative management of liver and spleen injuries in neurologically impaired adult patients. Arch Surg 131:309-315, 1996 12. Villalba MR, Howells GA, Lucas RJ, et al: Non-operative management of the adult ruptured spleen. Arch Surg 125:836-840, 1990 13. Keller MS, Vane DW: Management of pediatric blunt splenic injury: The comparison of pediatric and adult trauma surgeons. J Pediatr Surg 30:221-225, 1995 14. Hunt JP, Lentz CW, Cairns BA, et al: Management and outcome of splenic injury: The results of a five year state wide population-based study. Am Surg 62:911-917, 1996
Discussion D. Mooney (Boston, MA): The increase in nonoperative management appeared to occur to regional hospitals in 1987 before your educational program was instituted, and I was wondering if you could elaborate on that? Secondly, at your institution, my understanding is that
the adult surgeons manage either the 16 or 17 year olds. I wonder why you included patients treated at your institution by the adult surgeons in your data set instead of just purely the patients treated by pediatric surgeons. Could you comment on the number of patients admit-
BLUNT SPLENIC INJURIES: CHANGING PATTERNS OF CARE
ted to your facility with this injury who are transferred versus admitted directly from the scene? It may be that the more stable patients were transferred to a facility and the other patients were treated at the referring hospital. Data studies from the State of Vermont, when they talk about transfer to a pediatric trauma center, need to consider the increase from 1993 through 1999 and the number of patients from Vermont transferred to Dartmouth in your manipulations. Finally, you imply in your data that your institution is a Level 1 pediatric trauma center and was through this study period. Could you just comment on how much of the study time your institution was actually a College of Surgeons Level 1 pediatric trauma center? C. Frumiento (response): I believe the first question dealt with when we instituted our program in 1990 and we start to see an increase in 1987. Actually, our graph does not show a very significant increase before 1990 in the regional hospitals. Our educational programs I think played a role in this. But we would need a control group in which no educational programs were done to truly prove that they were the sole reason for this increase. And we do not make that claim that the educational programs were the sole reason for the increase. We believe that physicians working around the state do try to stay current on the literature, and we just provide an additional resource to give them information as to how to manage these children. Your second question had to do with adults. It sort of gets mixed, and it gets a little blended in our institution. We like to say that 17 years and younger go to the pediatric surgery service. However, there are many instances in which we have had 18 and 19 year olds in our service. It is a good point. And we could look back and probably remove a handful of patients from this study group if we look specifically at 17 and under. As for transferring patients who are more stable, that is another very good point. But you also will notice that the ISS at the pediatric trauma centers were actually higher. That does not necessarily mean they were less stable, just
989
potentially more injured, perhaps had a higher score of injury. So it sort of cuts both ways. Yes, they might have kept and operated on some patients that were more unstable, but they also might have transferred some of those patients to a tertiary care center. As far as the level 1 status of our institution, I believe it gained level 1 status in the early 1990s. To that end, it was still a tertiary referral center for injured children with pediatric surgeons on staff throughout the study period. As far as Dartmouth, you are absolutely right, some children were referred to Dartmouth over the study period. Our database included all Vermont citizens, if you will, so we actually do have the data for patients that were transferred to Dartmouth. And in the last 5 years there were a handful, about 6 to 8 patients that were transferred to Dartmouth for care. That would change that P value, would raise the P value somewhat on that comparison looking at treatment at regional hospitals versus pediatric trauma center, increasing the number of patients that were referred to a trauma center. However, it would still leave the trend intact. With more patients, I think it would become statistically significant. R. Pearl (Peoria, IL): Your 70% plus nonoperative rate at Children’s Hospital actually is a very low. Most published series in the mid 1980s were about 85%, in the 1990s it is over 90%, with splenic salvage rates with splenorrhaphy of 95% to 97%. I wonder, are your rates as low as they are because general surgeons are caring for them at your center, you are including 19-year-olds, or do you have a population with higher ISS and therefore sometimes you do splenectomies because, for example, they had a perforated duodenum? C. Frumiento (response): These are very good questions. Our rates in the study period 1992 through 1995 were actually consistently above 85%, in the last several years in the study were above 90% looking at our graph. Before that, the rates were somewhat on the low side. But again, you mentioned it yourself, the ISS were a little higher than some of the ISS quoted in these other studies.