ORIGINAL RESEARCH

Optimization of propofol dose shortensprocedural sedation time, prevents resedationand removes the requirement forpost-procedure physiologic monitoringAnthony Bell, Greg Treston, Robert Cardwell, W Jacobus Schabort and Dip ChandDepartment of Emergency Medicine, Redcliffe Hospital, Redcliffe, Queensland, Australia

Abstract

Objective: To examine the effect of propofol dosing (total dose and number of doses) on patientsedation time and likelihood of resedation.

Methods: This was a prospective, observational patient series in an urban district hospital ED with42 000 attendances per annum. Patients undergoing an emergent procedure requiringprocedural sedation were included. Titrated intravenous propofol was administeredaccording to departmental procedure. Standardized consent and data collection forms wereused. Time taken for the patient to become conversational after first administration wasrecorded and any resedation documented.

Results: Four hundred patients, undergoing 404 procedures, were enrolled for the period commenc-ing August 2004 until March 2006. The mean initial propofol bolus was 0.8 mg/kg (SD 0.6),and mean total propofol dose was 1.8 mg/kg (SD 1.0), comprising a mean of 2.3 (SD 2.1)doses of 15.8 mg (SD 11.4). Mean sedation time was 11.8 min (SD 6.9), and increasedsedation times were associated with higher total propofol dose and number of boluses(P < 0.0001). Resedation occurred in two patients (0.5%, 95% CI 0–1.6%).

Conclusion: Shorter sedation times are seen with lower doses of propofol. Patients do not need pro-longed post-procedure monitoring because the occurrence of spontaneous resedation asso-ciated with propofol use is a rare event. This has implications for patient flow and staffresource allocation in a busy ED.

Key words: anaesthesia recovery period, conscious sedation, emergency medicine, monitoring, physiologic,propofol.

Correspondence: Dr Anthony J Bell, Department of Emergency Medicine, Royal Brisbane and Women’s Hospital, Butterfield Street,Herston, Qld 4006, Australia. Email: anthony_bell@health.qld.gov.au

Anthony Bell, MB BS, FACEM, Staff Specialist; Greg Treston, BMedSci, MB BS, DTMH(Lon), D.IMC.RCS(Ed.), FACEM, FACRRM, StaffSpecialist; Robert Cardwell, MB BS, FACEM, Staff Specialist; W. Jacobus Schabort, MBChB, Registrar; Dip Chand, MB BS, Registrar.

doi: 10.1111/j.1742-6723.2007.01009.x Emergency Medicine Australasia (2007) 19, 411–417

© 2007 The AuthorsJournal compilation © 2007 Australasian College for Emergency Medicine and Australasian Society for Emergency Medicine

Introduction

Procedural sedation in the ED is commonly undertakenin order to perform emergency painful procedures, suchas orthopaedic manipulations and cardioversion. Proce-dural sedation requires considerable staff involvementand takes time. Current Australasian College for Emer-gency Medicine Guidelines require that two medicalattendants, one of whom should be a specialist oradvanced trainee, be present. Nursing staff are alsorequired, and the procedure must be performed in aresuscitation area. Physiologic monitoring is mandatedduring the procedure and extending into the recoveryphase.1

Propofol was chosen as the preferred agent for EDprocedural sedation because of its availability, rapidonset and short duration of action, suitability for titra-tion, sedative hypnotic, amnestic and anti-emetic pro-perties.2 The optimum dosing of propofol for proceduralsedation has yet to be defined with single bolus,3–5

titration,6–10 bolus and infusion,11,12 or infusion alone,13,14

all being reported. In general, 1 mg/kg initial bolus and0.5 mg/kg subsequent boluses are commonly reporteddosages for propofol administration in the ED proce-dural sedation setting.15–17 In children, initial doses of2 mg/kg initial bolus have been suggested18,19 We con-sidered that the dose should be reduced further toenhance the ED utility of propofol.

Anecdotally, post-procedure monitoring in the EDmight be prolonged. Traditional reliance on opiate andbenzodiazepine usage, combined with application ofpost-general anaesthetic recovery practices, has beenone standard of practice. The duration of sedation isknown to be less with propofol compared withmidazolam,3–5,12 and Coll-Vincent et al. noted that rese-dation was also less with propofol, compared with otheragents, when used for cardioversion.3

The aim of the present study was to provide accurateinformation about duration of sedation and resedationrate using propofol, and to make recommendations aboutthe need for post-procedure physiologic monitoring.

Methods

This was a prospective, observational patient series per-formed in a district ED. Ethics approval was obtained,and patients were enrolled based upon their need for abrief, painful procedure. Written consent was obtained,for the use of propofol and the procedure itself, after

explanation of specific risks of procedural sedation andanalgesia (PSA). Other information has been extractedand published from these same data on the need forfasting; therefore, readers are referred to Bell et al. forfull details of methodology and ASA definitions.20

The specific end points of interest in this part of thestudy were time to ‘conversational’ (TTC), defined asthe time taken from first dose of propofol until patientsable to give appropriate responses to questions, andwhether resedation did or did not occur.

Descriptive summary statistics are provided. In thepresent study, a multiple regression model for theTTC (the dependant variable) was used in order todetermine which patient variables were significant,independent of one another. Those were age (years),weight (kg), total dose of propofol (mg), total numberof doses propofol (mg) and opiate dose expressed asmorphine equivalents (where 0.1 mg/kg morphine wasassumed to be equipotent to 1 mg/kg fentanyl).Dummy variables were constructed for ASA class 2 or3 in relation to ASA class 1, and whether an opiatehad been received or not. These were considered clini-cally relevant factors when attempting to provide pro-cedural sedation in the ED. Decision Tools and STAT

Tools Suite (student version) was the statistical soft-ware used.21

With respect to sample size, using multiple regres-sion a rule of thumb for testing coefficients is to haveN � 104 + the number of independent variables or 111patients (http://www.statisticssolutions.com/Multiple-Regression.htm#samplesize, accessed 14 February2007). The sample size was increased to allow safetyprofiling of resedation events, which were perceived tobe rare. A sample of 350 patients provides an upper95% confidence interval for that event of <2%.22

Results

Overall, 404 sedation events were undertaken using pro-pofol in 400 patients. Two hundred and twelve (52.5%,95% CI 47.5–57.4) were male. Two hundred and eighty-nine (72.3%, 95% CI 67.6–76.6) were ASA class 1, 89(22.3%, 95% CI 18.3–26.6) were ASA class 2, and 22(5.4%, 95% CI 3.5–8.2) of sedation episodes occurred inASA class 3 patients. (Four records had incompleteASA information.) Two patients (0.5%, 95% CI 0.0–1.6)had failed hip relocations, and two (0.5%, 95% CI 0.0–1.6) re-presented after dislocating their shoulders atanother time.

A Bell et al.

412 © 2007 The AuthorsJournal compilation © 2007 Australasian College for Emergency Medicine and Australasian Society for Emergency Medicine

The mean initial propofol bolus was 0.8 mg/kg (SD0.6). The mean total propofol dose was 1.8 mg/kg (SD1.0). A mean of 2.3 (SD 2.1) doses of 15.8 mg (SD 11.4)was given. Table 1 provides a breakdown of number ofsubsequent boluses and mean total dosing according toinitial dose range. There was a positive correlation(r = 0.5) between the initial bolus and total propofoldose (mg/kg) (Fig. 1).

The age distribution of patients and procedure type,along with their mean initial bolus (mg/kg), mean totalpropofol dose (mg/kg) and mean number of subsequentboluses (with SD), are shown in Table 2. These datashow that the mean initial and total doses administeredwere reduced as patients aged. There exists somedegree of between procedure variation in the doses ofpropofol administered.

The time taken for patients to become conversa-tional increased with an increasing total dose of pro-pofol. Mean TTC and 95% confidence intervals arerepresented graphically in Figure 2. Numbers ofpatients in each dose range were 77, 175, 89, 31 and17, respectively. (TTC was not recorded in 15 of the

episodes.) The independent variables associated withincreasing duration of sedation were total dose (mg/kg) and number of boluses (P < 0.0001). Opiate analge-sia and ASA class were not related to sedation times(P > 0.05). The multiple regression data with P-valuesfor independent variables are presented in Table 3.The adjusted r2 value was 0.25, suggesting thatunmeasurable patient and pharmacodynamic factorsplay a role.

Resedation occurred in two patients (0.5%, 95% CI0–1.6). The first patient was a previously well 75-year-old, 75 kg man undergoing reduction of shoulderfracture/dislocation. He had received 7.5 mg morphinepre procedure and a total of 2.7 mg/kg propofol. Thesecond was a 37-year-old, 82 kg man undergoing seda-tion for a patellar dislocation. He had received 50 mg offentanyl and a total of 1.33 mg/kg propofol. Both wereready for discharge from the department in 25 and45 min, respectively. One small but interesting group of12 patients, in our sample, were conversant throughoutthe procedure. They received a lower total mean dose ofpropofol (0.9 mg/kg).

Table 1. Initial bolus, total propofol dose and number of boluses

Initial bolus range(mg/kg)

No. patient(%, 95% CI)

Mean no.boluses (SD)

Total propofol dosein mg/kg (SD)

<0.5 106 (26.0%, 95% CI 22.0–30.8) 3.1 (2.8) 1.4 (0.9)0.5–1.0 209 (51.7%, 95% CI 46.7–56.7) 2.3 (1.5) 1.7 (0.8)>1.0 88 (21.8%, 95% CI 17.9–26.1) 1.5 (1.4) 2.4 (1.1)

Missing dose data in one record.

Correlation Coefficient 0.54

0

1

2

3

4

5

6

7

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5

Initial bolus (mg/kg)

To

tal p

rop

ofo

l d

ose (

mg

/kg

)

Figure 1. Correlation between initial bolus dose (mg/kg) and total propofol dose (mg/kg).

Propofol dose and sedation time

413© 2007 The AuthorsJournal compilation © 2007 Australasian College for Emergency Medicine and Australasian Society for Emergency Medicine

Discussion

Emergency physicians perform procedural sedation aspart of their daily work. The use of intravenous anaes-thetic agents such as propofol has been shown to be safein a wide range of settings, including the ED.23,24 TheAmerican College of Emergency Physicians supportlevel B evidence for the use of propofol (in their clinicalpolicy document on procedural sedation), bringing itinto line with their recommendations for fentanyl/midazolam.25

All procedures should be performed in resuscitationareas with physiologic monitoring, by personnel whoare adequately trained to manage an airway.1,25 Whatconstitutes the optimum agent or combination of agentsfor achieving successful and safe PSA in the ED isdebated. Midazolam/opioid combinations have beenperceived to provide more predictable response andhave a favourable safety profile when compared withpropofol, owing to the potential of propofol to inducedose-related deep sedation.26

Reducing the time a patient is sedated potentiallyreduces the time a patient is at risk of airway compro-mise or respiratory depression. Sedation times areshorter with propofol, and reported respiratory compli-cation rates for propofol are equivalent to midazolamalone,12 midazolam � flumazenil or etomidate,3 andmidazolam/fentanyl.5 Godambe et al. describe moredesaturation, but at doses greatly in excess of our data:4.5 versus 1.8 mg/kg total propofol dose.4 ‘Waiting out’respiratory depression in a propofol sedated but pre-oxygenated patient is reasonable.17 Our data supportshorter sedation times when lower total propofol dosesand fewer boluses are given. The variation in initialdose and bolus doses, in addition to the inclusion ofoutliers who received many small boluses, makes inde-pendence possible. Although only 25% of variation in

Table 2. Age and procedure-related dosing profiles

Mean initialdose in mg/kg (SD)

Mean totaldose in mg/kg (SD)

Mean no.subsequent boluses (SD)

Age�15 years (n = 112) 1.0 (0.7) 2.3 (1.2) 2.6 (1.7)16–64 years (n = 203) 0.8 (0.5) 1.7 (1.0) 2.4 (2.4)�65 years (n = 83) 0.6 (0.3) 1.1 (0.5) 1.7 (1.4)

Procedure typeJoint relocation

Shoulder (n = 94) 0.8 (0.4) 1.4 (0.7) 1.8 (1.7)Elbow (n = 8) 0.8 (0.3) 1.3 (0.6) 1.0 (1.0)Ankle (n = 6) 0.7 (0.4) 1.4 (0.6) 2.7 (2.2)Hip (n = 20) 0.6 (0.3) 1.3 (0.6) 2.2 (1.4)Patellar (n = 4) 0.7 (0.2) 1.1 (0.4) 1.3 (1.3)Total (n = 132) 0.8 (0.4) 1.4 (0.7) 1.8 (1.7)

Closed reductionUpper limb (n = 180) 0.8 (0.6) 1.9 (1.0) 2.4 (1.6)Lower limb (n = 23) 0.8 (0.7) 2.0 (1.5) 2.7 (2.1)Total (n = 203) 0.8 (0.6) 1.9 (1.1) 2.5 (1.7)

Incision and drainage (n = 39) 1.0 (0.7) 2.3 (1.1) 2.9 (2.3)Cardioversion (n = 6) 0.7 (0.3) 1.1 (0.5) 1.3 (0.8)Other (n = 20) 1.0 (0.8) 2.4 (1.5) 3.6 (5.0)

Missing age data in 6 records; missing indication data in 4 records.

0

5

10TT

C

Total propofol dose (mg/kg)

15

20

25

<1 1–2 2–3 3–4 4+

Figure 2. Line graph of ‘time to conversational’ (TTC) andincreasing total propofol dose. –�– mean TTC.

A Bell et al.

414 © 2007 The AuthorsJournal compilation © 2007 Australasian College for Emergency Medicine and Australasian Society for Emergency Medicine

sedation time can be explained, it has clinical relevancebecause this is the thing we have control over. Further-more, reducing the initial propofol bolus contributes toan overall reduction in the total dose given. Patient-controlled sedation techniques have the potential tofurther reduce the dose of propofol used in PSA.

Our findings are consistent with that of Coll-Vincentet al. and indicate that resedation following propofol israre.3 Such clinical correlation with the known pharma-cologic profile of propofol is reassuring and, as such,once the patient can talk, nursing staff have an ‘end-point’ for the cessation of physiologic monitoring, con-fident that patients are unlikely to develop any adverseevents at this time. This stance is supported by a recentClinical Practice Advisory from the USA.27 At a depart-mental level, patient flows within the ED might beimproved, patient movement to X-ray is facilitated, andbusy monitored areas are free for the next patient.Nursing allocation can be tailored to these less intensivepost-procedure monitoring requirements.

Depth of sedation was not objectively scored in ourstudy. The exact characteristics of respiratory and/orairway reflex depression in relation to depth of sedationare not well defined.28 However, careful dose titrationand subjective evaluation of patient responsivenessthroughout the procedure is recommended.23,26 Our data,along with that of Burton et al., show that higher totaldoses tend to be used in children compared withadults.29 Frazee et al. describe a similarly conservativeapproach, to the investigators, utilizing dose reductionin those over 65 years of age.11 Other authors have uti-lized the Observer’s Assessment of Alertness/SedationScale, validated for midazolam.30 Miner and colleaguesdescribe less respiratory depression when BispectralElectroencephalographic Analysis was used in conjunc-tion with propofol.31,32 One small but interesting groupof 12 patients, in our sample, were conversant through-out the procedure.

Potential limitations of the study were that patientexclusion criteria were not specifically defined. It wasleft to the discretion of the treating emergency physi-cian. The initial intention was to sedate only ASA classI and II patients. There were a small number of ASAclass III patients sedated in this cohort out of necessity.Our convenience sample might have introduced selec-tion bias, as no effort was made to record the number ofpatients receiving agents other than propofol for EDsedation. However, it was uncommon that agents otherthan propofol were used in teenagers and adults in thisdepartment. It was the intention that propofol be admin-istered according to the dose provided in the protocol,but individual physicians did not always comply withthe suggested weight-based initial bolus dose, hence thevariation in initial dose. These patients receiving higherinitial boluses have contributed to our understanding oftotal dose on sedation times. When performing the mul-tiple regression, independent variables were chosenbased on their perceived clinical utility. Intrinsic patientfactors are certainly an issue in prolonged sedationtimes, and the low adjusted r2 value reflects this.

Conclusion

We believe that procedures can be successfully per-formed using lesser overall doses of propofol than havebeen reported previously. The present study demon-strates that in increased doses, propofol use is associatedwith longer sedation times. As such, starting low andtitrating up based on the feedback gained through con-stant re-assessment of the patient’s responsiveness is theauthors’ recommendation. This will minimize the dura-tion of sedation and the period at risk for adverse events.Resedation occurs infrequently ,so once a patient is con-versational, it is safe to take them off their monitoring.

Table 3. Multivariate analysis: time to conversational as a function of measurable patient variables

Regression table Coefficient Lower limit Upper limit P-value

Constant 3.5 1.5 5.4 0.0005Age (years) 0.0 0.0 0.1 0.33Total number of boluses 1.1 0.8 1.4 <0.0001Total propofol dose (mg/kg) 1.7 0.9 2.4 <0.0001Total opiate dose (mg/kg) 6.0 -2.3 14.4 0.16ASA ‘class 2’ 0.0 -1.8 1.8 1.00ASA ‘class 3’ 2.2 -0.9 5.2 0.16Received opiate 0.7 -1.0 2.5 0.42

Propofol dose and sedation time

415© 2007 The AuthorsJournal compilation © 2007 Australasian College for Emergency Medicine and Australasian Society for Emergency Medicine

Potential exists for reduced occupancy of beds in resus-citation areas and might facilitate patient flows.

Acknowledgements

Dr Mark Stickley, MB BS MPH, for statistical advice.Staff, patients and medical students of Redcliffe DistrictHospital ED.

Author contributions

AB, GT and RC conceived the study and submitted theethics application. AB performed the statistical analysisand presented the data. All authors recruited patients,ensured completeness of data forms, were involved instaff education, and were responsible for administeringthe database. AB takes responsibility for the paper as awhole.

Competing interests

None declared.

Accepted 8 June 2007

References

1. Australasian College for Emergency Medicine. Statement onClinical Principles for Procedural Sedation,2003. [Cited 21 August2006.] Available from URL: http://www.acem.org.au/media/policies_and_guidelines/P29_ANZCA.pdf

2. Bahn EL, Holt KR. Procedural sedation and analgesia: a reviewand new concepts. Emerg. Med. Clin. N. Am. 2005; 23: 503–17.

3. Coll-Vincent B, Sala X, Fernandez C et al. Sedation for cardiover-sion in the emergency department: analysis of effectiveness infour protocols. Ann. Emerg. Med. 2003; 42: 767–72.

4. Godambe SA, Elliot V, Matheny D, Pershad J. Comparison ofpropofol/fentanyl versus ketamine/midazolam for brief ortho-paedic procedural sedation in a paediatric emergency depart-ment. Pediatric 2003; 112: 116–23.

5. Taylor D, O’Brien D, Ritchie P, Pasco J, Cameron PA. Propofolversus midazolam/fentanyl for reduction of anterior shoulderdislocation. Acad. Emerg. Med. 2005; 12: 13–19.

6. Guenther Skokan E, Pribble C, Bassett KE, Nelson DS. Use ofpropofol sedation in a pediatric emergency department: a pro-spective study. Clin. Pediatr. 2001; 40: 663–71.

7. Miner JR, Biros MH, Krieg S, Johnson C, Heegaard W, PlummerD. Randomised clinical trial of propofol versus methohexital forprocedural sedation during fracture and dislocation reduction inthe emergency department. Acad. Emerg. Med. 2003; 10: 931–7.

8. Miner JR, Heegaard W, Plummer D. End tidal carbon dioxidemonitoring during procedural sedation. Acad. Emerg. Med. 2002;9: 275–80.

9. Guenther E, Pribble CG, Junkins EP, Kadish HA, Bassett KE,Nelson DS. Propofol sedation by emergency physicians for elec-tive pediatric outpatient procedures. Ann. Emerg. Med. 2003; 42:783–91.

10. Bassett KE, Anderson JL, Pribble CG, Guenther E. Propofol forprocedural sedation in children in the emergency department.Ann. Emerg. Med. 2003; 42: 773–82.

11. Frazee BW, Park RS, Lowery D, Baire M. Propofol for deepprocedural sedation in the ED. Am. J. Emerg. Med. 2005; 23:190–5.

12. Havel CJ, Strait RT, Hennes H. A clinical trial of propofol versusmidazolam for procedural sedation in a pediatric emergencydepartment. Acad. Emerg. Med. 1999; 6: 989–96.

13. Swanson ER, Seaberg DC, Mathias S. The use of propofol forsedation in the emergency department. Acad. Emerg. Med. 1996;3: 234–8.

14. Pershad J, Godambe SA. Propofol for procedural sedationin the emergency department. J. Emerg. Med. 2004; 27: 11–14.

15. Symington L, Thakore S. A review of the use of propofol forprocedural sedation in the emergency department. Emerg. Med.J. 2006; 23: 89–93.

16. Green SM, Krauss B. Propofol in emergency: pushing the seda-tion frontier. Ann. Emerg. Med. 2003; 42: 792–7.

17. Krauss B, Green S. Procedural sedation and analgesia in children.Lancet 2006; 367: 766–80.

18. Saccheti A, Cravero J. Sedation in the emergency department.Pediatr. Ann. 2005; 34: 617–22.

19. Barnett P. Propofol for pediatric sedation. Pediatr. Emerg. Care2005; 21: 111–14.

20. Bell A, Treston G, McNabb C, Monypenny K, Cardwell R. Pro-filing adverse respiratory events and vomiting when using pro-pofol for emergency department procedural sedation. Emerg.Med, Australas. 2007; 19: 405–410 (in press).

21. Thomson Learning.The Decision Tools and Stat ToolsSuite,2006. [Cited 21 June 2006.] Available from URL: http://www.palisade.com

22. Jovanovic BD, Zalenski RJ. Safety evaluation and confidenceintervals when the number of observed events is small or zero.Ann. Emerg Med. 1997; 30: 301–6.

23. Ducharme J. Propofol in the emergency department: anotherinterpretation of the evidence. J. Can. Assoc. Emerg. Phys. 2001;3: 311–12.

24. Jackson R. Use of propofol for procedural sedation in the emer-gency department. Emerg. Med. J. 2000; 5: 378–9.

25. American College of Emergency Physicians. Clinical policy: pro-cedural sedation and analgesia in the emergency department.Ann. Emerg. Med. 2005; 45: 177–96.

26. Green SM. Propofol in the emergency department – not yet readyfor prime time. Acad. Emerg. Med. 1999; 6: 975–8.

27. Miner JR, Burton JH. Clinical practice advisory: Emergencydepartment procedural sedation with propofol. Ann. Emerg.Med. 2007; 50: 182–7.

A Bell et al.

416 © 2007 The AuthorsJournal compilation © 2007 Australasian College for Emergency Medicine and Australasian Society for Emergency Medicine

28. Green SM, Roback MG, Miner JR, Burton JH, Krauss B. Fastingand emergency department procedural sedation and analgesia: aconsensus-based clinical practice advisory. Ann. Emerg. Med.2007; 49: 454–61.

29. Burton JH, Miner JR, Shipley ER, Strout TD, Becker C, Thode HC.Propofol for emergency department procedural sedation andanalgesia: a tale of three centers. Acad. Emerg. Med. 2006; 13:24–30.

30. Chernik DA, Gillings D, Laine H et al. Validity and reliability ofthe observer’s assessment of alertness/sedation scale: study with

intravenous midazolam. J. Clin. Psychopharmacol. 1990; 10: 244–51.

31. Miner J, Biros MH, Seigel T, Ross K. The utility of the bispectralindex with propofol in the emergency department. Acad. Emerg.Med. 2005; 12: 190–6.

32. Miner JR, Biros MH, Heegaard W, Plummer D. Bispectral elec-troencephalographic analysis of patients undergoing proceduralsedation in the emergency department. Acad. Emerg. Med. 2003;10: 638–43.

Propofol dose and sedation time

417© 2007 The AuthorsJournal compilation © 2007 Australasian College for Emergency Medicine and Australasian Society for Emergency Medicine