On-Scene Intravenous Line Insertion Adversely Impacts Prehospital Time in Rural Vehicular Trauma RICHARD P. GONZALEZ, M.D.,*t GLENN R. CUMMINGS, R.N.,»t HERBERT A. PHELAN, M.D.,* MADHURI S. MULEKAR, PH.D.,+ CHARLES B. RODNING, M.D.*
From the Departments of*Surgery, f Statistics and Mathematics, and jiCenter for the Study of Rural Vehicular Trauma, University of South Alabama, Mobile, Alabama; and the National Highway Traffic Safety Administration, Washington, DC Fatality rates from rural vehicular trauma are almost double those found in urban settings. Increased emergency medical services (EMS) prehospital time has been implicated as one of the causative factors for higher rural fatality rates. Advanced Trauma Life Support guidelines suggest scene time should not be extended to insert an intravenous catheter (IV). The purpose of this study was to assess the association between intravenous line placement and motor vehicle crash (MVC) scene time in rural and urban settings. An imputational methodology using the National Highway Traffic Safety Administration Crash Outcome Data Evaluation System permitted linkage of data from police motor vehicle crash and EMS records. Intergraph GeoMedia software permitted this linked data to be plotted on digital maps for segregation into rural and urban groups. MVCs were defined as rural or urban by location of the accident using the U.S. Bureau of Census Criteria. Linked data were analyzed to assess for EMS tjme on-scene, on-scene IV insertion, on-scene IV insertion attempts, and patient mortality. Over a 2-year period from January 2001 through December 2002, data were collected from Alabama EMS patient care reports (PCRs) and police crash reports. A total of 45,763 police crash reports were linked to EMS PCRs. Of these linked crash records, 34,341 (75%) and 11,422 (25%) were injured in rural and urban settings, respectively. Six hundred eleven (1.78%) mortalities occurred in rural settings and 103 (0.90%) in urban settings (P < 0.005). There were 6,273 (18.3%) on-scene IV insertions in the rural setting and 1,290 (11.3%) in the urban setting (P < 0.005). Mean EMS time on-scene when single IV insertion attempts occurred was 16.9 minutes in the rural setting and 14.5 minutes in the urban setting (P < 0.0001). When two attempts of on-scene IV insertion were made, mean EMS time on-scene in the rural setting (n = 891 [2.6%]) was 18.4 minutes and 15.7 minutes in the urban setting (n = 142 [1.2%; P < 0.005). Excluding dead on-scene patients, mean EMS time on-scene when mortalities occurred in rural and urban settings was 18.9 minutes and 10.8 minutes, respectively (P < 0.005). On-scene IV insertion occurred with significantly greater frequency in rural than urban settings. This incurs greater EMS time on-scene and prehospital time that may be associated with increased vehicular fatality rates in rural settings.
HE PURPOSE OF Advanced Life Support (ALS) in trauma patients is stabilization of the patient before hospital arrival. Basic tenets of ALS are airway
T
Presented during Poster Grand Rounds at the Annual Scientific
and oxygénation maintenance, spinal stabilization, and intravenous (IV) access for fluid replacement and administration of medications. Although these basic tenets have been the cornerstone of prehospital trauma care for many years, the efficacy of ALS prehospital ^^j-g ¡^ reducing trauma-related mortality has been
Meeting and Postgraduate Course Program, Southeastern Surgical
Congress, Savannah, Georgia, Februa^ 10-13, 2007. This work was performed under a cooperative agreement with the U.S. Department of Transportation/National Highway Safety Traffic Administration (USDOT/NHTSA), Grant #DTHN 22-00H-05285. Views expressed are those of the authors and do not represent the views of the sponsors or NHTSA. Address correspondence and reprint requests to Richard P. Gonzalez, M.D., 2451 Fillingim Street, Mobile, AL 36617. E-mail:
[email protected].
.
J I S T - - I
•
ATC-.
r ^
questioned. •-» Intuitively, on-site ALS care for trauma patients should reduce mortality by preventing hemodynamic deterioration; however, many clinicians feel ALS prehospital care can contribute to patient mortaljj ^y delaying time tO definitive in-patient hospital ~ r u i--i- • l ^^^re. Supporters of the on-SCene Stabilization approach claim intravenous therapy benefits mortality by timely improvement in oxygen delivery to vital organs. Pro-
1083
1084
THE AMERICAN SURGEON
ponents of the "scoop and run" approach claim attempts at on-scene IV insertion should be avoided to reduce prehospital time. Studies that support on-scene ALS care report that procedures such as prehospital intubations and IV line insertions have failed to show a direct reduction in mortality.^"'^ Studies that support the "scoop and run" methodology purport lower mortality risk. '' ^' ^- ^' ^ Most of these series are small, uncontrolled, and retrospective and therefore lack the power to draw definitive conclusions. Bickell and colleagues performed a randomized, prospective study comparing early versus delayed IV fluid therapy and concluded that delayed fluid replacement improved mortality.'^ The Committee on Trauma of the American College of Surgeons Resources for the Optimal Care of the Injured Patient guidelines for on-scene resuscitation of the severely injured patient suggest on-scene resuscitative maneuvers should be limited to establishment of an airway, provision of ventilation, hemorrhage control, stabilization of fractures, and immobilization of the entire spine.^^ These guidelines also suggest that IV access may be established en route to the hospital and scene time should not be extended to start an IV line. Increased emergency medical services (EMS) prehospital time has been implicated as a contributing factor to greater mortality rates from vehicular trauma in rural areas. The purpose of this study was to assess the frequency of EMS on-scene IV insertion and the association of IV insertion to EMS on-scene time in rural and urban settings. Methods Data were retrospectively collected from EMS Patient Care Reports (117,605) (EMS-PCRs) and Police Accident Reports (274,175 PARS and 741,007 occupant records) from the State of Alabama during the 2-year period from January 2001 through December 2002. Linkage of crash data to EMS data was consummated using Crash Outcome Data Evaluation System (CODES2000, Strategic Matching, Inc., Morrisonville, NY) software. Linking data sources together provide traffic safety researchers with a broader picture of the crash scenario plus opportunities for measuring outcomes related to traffic safety improvement efforts. CODES2000 software use probabilistic linkage techniques to join data sets by the use of indirect identifiers that differentiate among the events and the persons involved. Crash characteristics (i.e., time, location, object struck), person characteristics (i.e., age, sex), and vehicle characteristics (i.e., type of vehicle) that are common across data sets can be used to link person level records. CODES2000 software also uses
November 2008
Vol. 74
techniques to impute missing links and missing values based on available known state-specific data in the aggregate. The linked and imputed data are checked for validity and consistency to ensure that the data were not biased and were representative of the population. A 95% CI and 12 joint specifications were used to link data from PARs and EMS-PCRs. The 12 joint specifications used to link records were: sex, age, race, position in vehicle, county where motor vehicle crash (MVC) occurred, date of MVC, time of MVC, weather conditions, vehicular speed greater than 55 mph, safety equipment used, airbag deployment, and injuries sustained. All patients entered in the study included patients who were admitted to a medical facility, who died in an emergency department, or were transferred to another medical facility for a higher level of care. Patients who were dead at the scene and patients who required prolonged extrications (greater than 20 minutes) were excluded from this analysis. Geographic Information System (GIS) software (Intergraph Geomedia, Madison, AL) was used to differentiate between rural and urban areas as defined by the U.S. Bureau of Census.^' The Bureau of Census defined a densely settled territory of 50,000 or more inhabitants with concentration exceeding 1,000 persons per square mile as an urban area. Adjacent cities, towns, and villages with populations exceeding 2,500 were also considered urban. All other areas that did not conform to the urban definition were considered rural. The GIS software allowed for differentiation between rural and urban areas and identified locations for all MVCs based on city and county codes extracted from the PARs. Linked EMS-PCRs and PARs were analyzed for evaluation for on-scene location, EMS time on-scene, on-scene IV insertion, on-scene IV insertion attempts, and patient mortality. Fisher's exact test was used to assess the association between frequency of IV insertion and rural or urban setting. Student's / test was used to compare mean scene times between rural and urban settings. The significance level used for each test P < 0.05. Results
A total of 45,763 PARs were linked to EMS-PCRs. Thirty-four thousand three hundred forty-one (75%) patients were linked from rural areas and 11,422 (25%) patients were linked from urban areas. Six hundred eleven (1.8%) mortalities occurred in rural areas and 103 (0.9%) occurred in urban areas. Excluding patients who were dead on-scene (DOS) and patient extrications greater than 20 minutes, 18.3 per cent of the patients from a rural area underwent
No. 11
PREHOSPITAL TIME IN RURAL VEHICULAR TRAUMA
on-scene IV insertion and 11.3 per cent of patients from urban areas underwent on-scene IV insertion ( P < 0.001) (Table 1). Excluding patients who were DOS and patient extrications greater than 20 minutes, mean EMS time on-scene in rural areas was 16.9 minutes (SD, 9.9) when IV insertions occurred and 14.1 minutes (SD, 9.7) when IV insertions were not performed {P < 0.001). In urban areas, mean EMS time on-scene was 14.5 minutes (SD, 6.7) when IVs were inserted and 12.0 minutes (SD, 6.1) when IVs were not inserted (P < 0.001) (Table 2). When IVs were inserted onscene, mean EMS time on-scene was significantly greater in rural crasb scenes than urban crash scenes; however, EMS on-scene time was also greater at rural crash scenes than urban crash scenes when IVs were not inserted (Table 3). Increasing EMS on-scene IV insertion attempts was associated with increasing EMS time on-scene for both rural and urban areas. As IV insertion attempts increased, EMS time on-scene was significantly greater in rural areas than urban areas (Table 4). Excluding patients who were DOS and prolonged extrications, mean EMS time on-scene when mortalities occurred in rural and urban areas was 18.9 minutes and 10.8 minutes, respectively (P < 0.005). Discussion Proponents of on-scene IV insertion claim IV fluid replacement improves hemodynamic status of patients and decreases the incidence of hemorrhagic shock.^"'^ Those who support on-scene IV insertion claim that EMS personnel who are experienced with IV insertion can do so with minimal delays; however, no trial has demonstrated improved patient survival with this procedure performed on-scene. Supporters of the "scoop and run" approach have shown that this method of patient retrieval is associated with reduced risk of patient mortality.'"^ The American College of Surgeons Committee on Trauma supports insertion of IVs en route to the hospital and discourages IV insertion at the scene if IV insertion extends scene time.^^ Trunkey claims that IV fluid replacement is not clinically effective because the amount of blood lost as a result of severe bleeding cannot be replaced by IV fluid administered at the injury scene.^^ Moreover, Trunkey states that time TABLE
1. Rural vs Urban Intravenous Line Insertions Intravenous Line Insertions (%)
Rural Urban
6,273(18.3) 1,290(11.3) P < 0.0001
Gonzalez et al.
1085
2. Scene Times and Intravenous Line (IV) Insertion vs Location
TABLE
Rural Urban
IV insertion No IV insertion IV insertion No IV insertion
Number
Mean Scene Time (minutes)
5,693 15,731 1,057 5,279
16.9 14.1 14.5 12.0
P< 0.0001 P < 0.0001
TABLE 3. Scene Times and Location vs Intravenous Line (IV) Insertion
IV insertion No IV insertion
Rural Urban Rural Urban
Number
Mean Scene Time (minutes)
5,693 1,057 15,731 5,279
16.9 14.5 14.1 12.0
f < 0.0001 P < 0.0001
spent at the scene for IV insertion increases prehospital delay and may increase patient mortality. Severely injured patients who are bleeding require timely surgical intervention and bleeding control to optimize survival. Sampalis et al. recently reported on 44 preventable deaths occurring in a cohort of 360 patients with major trauma.^-' These authors concluded that 34 patients who required direct transport to a Level I trauma center showed significant prehospital delays and high rates of inappropriate IV line insertion. More recently, Sampalis et al. has shown in an observational study of 217 patients that had on-site IV fluid replacement was associated with increased mortality risk and this association was exacerbated by increased prehospital times.^'^ Our data has shown that mortality rates from MVCs are twice that found in rural areas than urban areas in the state of Alabama. Furthermore, when patients who were DOS and prolonged extrication patients were excluded, mean EMS scene time in rural areas was 75 per cent higher than in urban areas (18.9 minutes vs 10.8 minutes; P < 0.005). Associated with these findings are rural on-scene IV insertion rates that are 73 per cent higher than in urban areas (Table 1). The American College of Surgeons Committee on Trauma suggest that IV insertion should not prolong scene time; however, based on our data, there appears to be prolongation of scene time resulting from IV insertion in both rural and urban settings (Table 2). In both urban and rural areas, insertion of IV lines was associated with significantly longer scene times. This concurs with Trunkey's^^ assertion and Gold's^^ assertion that time spent on-scene for initiation of IV lines will
1086
TABLE
THE AMERICAN SURGEON
November 2008
Vol. 74
4. Scene Times and Number of Intravenous Line (IV) Insertions vs Location Rural
Number of IV Attempts None One Two
Three
Urban Number
Mean Seene Time (minutes)
15,731 4,561
14.1 16.5 18.4 18.9
891 241
P < 0.0001
increase prehospital delays and increase risk of mortality. Based on our data, it is clear that on-scene IV insertion occurs at significantly greater rates in rural settings and rural mortality rates resulting from MVCs are double those found in urban areas. Furthermore, when IV lines are inserted on-scene, rural scene times are significantly longer than urban scene times (Table 3). When mortalities occurred (excluding patients who were DOS and those undergoing prolonged extrication), scene times were 75 per cent greater in rural areas than urban areas (18.9 minutes vs 10.8 minutes; P < 0.005). In rural areas, IVs were inserted with significantly greater frequency and, when inserted, took longer to achieve successful insertion. Based on our data, it would be incorrect to conclude on-scene IV line insertion directly causes increased mortality rates in rural areas; however, it could be concluded that increased scene times and increased IV insertion rates in rural areas does not improve patient survival and may negatively impact rural mortality rates. A factor that appears to contribute to increased scene times in rural areas when IV lines are attempted is persistence with on-scene IV insertion. Second and third IV insertion attempts occurred at significantly higher rates in rural areas and when increased attempts occurred, on-scene time increased significantly (Table 4). Multiple on-scene insertion attempts are in direct conflict with American College of Surgeons guidelines that IV line insertion should not prolong scene time. Several authors have supported en route IV insertion as a method of IV line insertion that has high insertion success rates without adversely affecting scene time.^''"^^ In a prospective study, O'Gorman et al. found that en route IV insertion had higher success rates than on-scene IV insertion.^'^ Their study reported a higher en route IV insertion success rate in hypotensive patients and concluded that even if delay at the scene is minimal, it is not possible to justify any delay, because IV lines can be successfully instituted en route. Similarly, Slovis et al. reported an overall 92 per cent success rate for IV insertion on trauma patients in a moving ambulance with higher success rates in hypotensive patients.^^ These authors concluded
Number
Mean Seene Time (minutes)
5,279
12.0 14.2 15.8
876 142 39
16.7
P < 0.0001 P < 0.0001 P < 0.0001
P < 0.0001
that prompt transport of unstable patients should not be delayed solely to obtain IV access. Because the mortality rates reported here are the only clinical gauge available from a statewide repository, we can only draw the conclusion that increased scene times and IV insertion rates may contribute to increased rural mortalities. The combination of factors leading to motor vehicle deaths is numerous. Speed, vehicle type, use of safety devices, position in the vehicle, alcohol or drugs, pre-existing medical conditions, prehospital EMS care, and trauma system availability are just a few factors that must be considered while investigating and identifying areas that can be improved on to reduce vehicular mortality rates. The results of this study support conclusions drawn by Trunkey^^ and Sampalis^'* that prehospital delays associated with IV insertion contribute to mortality rates. Considered separately, increased scene time associated with IV insertion may not be clinically significant with regard to mortality rates. However, when bundled with other criteria such as prolonged EMS response times, inappropriate on-scene procedures, lack of EMS procedure proficiency, and increased travel times from the scene, the cumulative effect can increase prehospital times that affect clinical outcome. Mortality rates from rural vehicular trauma are twice the rates found in their urban counterparts. Onscene IV insertion occurs at significantly higher frequency in rural areas than urban areas. This incurs greater on-scene times in rural areas that may negatively impact patient survival. REFERENCES 1. Ivatury RR, Nallathambi MN, Roberge RJ, et al. Penetrating thoracic injuries: In-field stabilization vs prompt transport. J Trauma 1987;27:1066-73. 2. Tsai A, Kallsen G. Epidemiology of pédiatrie prehospital care. Ann Emerg Med 1987; 16:284-92. 3. Fielder MD, Jones LM, Miller DF, Finley RK. A correlation of response time and results of abdominal gunshot wounds. Arch Surg 1986;121:902^. 4. Mattox KI. Prehospital care of the patient with an injured chest. Surg Clin North Am 1989;69:21-9. 5. Pepe PE, Stewart RD, Copass MK. Prehospital management of trauma: A tale of three cities. Ann Emerg Med 1986; 15: 1484-90.
No. 11
PREHOSPITAL TIME IN RURAL VEHICULAR TRAUMA
6. Smith JP, Boda BI, Hill AS, Frey CF. Prehospital stabilization of critically injured patients: a failed concept. J Trauma 1985; 25:65-70. 7. Cayten CG, Longmore W, Kuehl A, et al. Basic life support vs advanced life support for urban trauma. J Trauma 1984;24:651. 8. Gervin AS, Fischer RP. The importance of prompt transport in salvage of patients with penetrating heart wounds. J Trauma 1982522:443-8. 9. Potter D, Goldstein G, Fung SC. A controlled trial of prehospital advanced life support in trauma. Ann Emerg Med 1988; 17:582-8. 10. Jacobs LM, Sinclair A, Beiser A, D'Agostino RB. Prehospital advanced life support: benefits in trauma. J Trauma 1984;24: 8-13. 11. Honigman B, Rohweder K, Moore EE, et al. Prehospital advanced life support for penetrating cardiac wounds. Ann Emerg Med 199O;I9:145-5O. 12. Reines HD, Bartlett RL, Chudy NE, et al. Advanced life support appropriate for victims of motor vehicle accidents: The South Carolina highway trauma project. J Trauma 1988;28: 563-70. 13. Cwinn AA, Pons PT. On-scene management of trauma patients by paramedics. Ann Emerg Med 1988;17:189-90. 14. Aprahamian C, Thompson BM, Towne JB, Darin JC. The effect of a paramedic system on mortality of major open intraabdominal vascular trauma. J Trauma 1983;23:687-90. 15. Hedges JR, Sacco WJ, Champion HR. An analysis of prehospital care of blunt trauma. J Trauma 1982;22:989-93. 16. Fortner GS, Oreskovich MR, Copass MK, et al. The effects of prehospital trauma care on survival from a 50-meter fall. J Trauma 1983;23:976-81. 17. Pons PT, Honigman B, Moore EE, et al. Prehospital advanced trauma life support for critical penetrating wounds to the thorax and abdomen. J Trauma l985;25:828-32. 18. Cwinn AA, Pons PT, Moore EE, et al. Prehospital advanced
Gonzalez et al.
1087
trauma life support for critical blunt trauma victims. Ann Emerg Med 1987; 16:399^03. 19. Bickell WH, Wall MJ, Pepe PE, et al. Immediate versus delayed fluid resuscitation for hypotensive patients with penetrating torso injuries. N Engl J Med 1994;331:1105-9. 20. Committee on Trauma. Resources for the Optimal Care of the Injured Patient. Chicago, IL: American College of Surgeons; 1999:13-7. 21. US General Accounting Office. Rural Development: Profile of Rural Areas. Fact Sheet for Congressional Requestors GAO/ RECD-93-40FS. Washington, DC: US General Accounting Office; 1993:26-30. 22. Trunkey DD. Is ALS necessary for pre-hospital trauma care? J Trauma 1984;24:86-7. 23. Sampalis JS, Boukas S, Lavoie A, et al. Preventable death evaluation of the appropriateness of the on-site trauma care provided by Urgences-Sante physicians. J Trauma 1995;39:IO29-35. 24. Sampalis JS, Tamim H, Denis R, et al. Ineffectiveness of on-site intravenous lines: is prehospital time the culprit? J Trauma 1997;43:608-17. 25. Trunkey DD. Trauma. Sei Am 1983;249:28-35. 26. Gold CR. Prehospital advanced life support vs 'scoop and run' in trauma management. Ann Emerg Med 1987; 16:797-801. 27. O'Gorman M, Trabulsy P, Pilcher DB. Zero-time prehospital IV. J Trauma 1989;29:84-6. 28. Slovis CM, Herr EW, Londorf D, et al. Success rates for initiation of intravenous therapy en route by prehospital care providers. Am J Emerg Med 1990;8:305-7. 29. Pace SA, Fuller FP, Dahlgren TJ. Paramedic decisions with placement of out-of-hospital intravenous lines. Am J Emerg Med 1999; 17:544-7. 30. Jones SE, Nesper TP, Alcouloumre E. Prehospital intravenous line placement: a prospective study. Ann Emerg Med 1989; 18:244-6.
