ORIGINAL ARTICLE

Local Anesthesia Combined With Sedation
Compared With General Anesthesia for
Ambulatory Operative Hysteroscopy:
A Randomized Study
Lone Dragnes Brix, MHSc, RN, Theis Muncholm Thillemann, MD, PhD,
Lone Nikolajsen, MD, DMSc

Purpose: The purpose of this study was to compare two anesthetic tech-
niques for postoperative pain after ambulatory operative hysteroscopy.
Design: A randomized trial.
Methods: Women (N 5 153) scheduled for ambulatory operative
hysteroscopy were assigned to receive either paracervical local anesthesia
combined with sedation (group LA 1 S; n 5 76) or general anesthesia
(group GA; n 5 77). Primary outcome was the worst pain intensity score
in the postanesthesia care unit (PACU) rated by the patients on a
numerical rating scale.
Finding: Data from 144 patients were available for analysis (LA 1 S:
n 5 69; GA: n 5 75). There were no significant differences in worst
pain intensity between groups in the PACU (P 5 .13) or after discharge
from PACU (P 5 .40). In group LA 1 S, fewer patients received treatment
with intravenous fentanyl intraoperatively (P , .01) and time until
discharge from PACU was shorter (P , .01). More patients in group
LA 1 S experienced vomiting after discharge (P , .05).
Conclusions: Local anesthesia with sedation can be recommended as a
first choice anesthetic technique for operative ambulatory hysteroscopy.
Keywords: postoperative pain, anesthesia, ambulatory surgery, opera-
tive hysteroscopy.
Ó 2016 by American Society of PeriAnesthesia Nurses

HYSTEROSCOPY PROCEDURES SUCH as poly- care providers, such as a decrease in complication
pectomy, endometrial ablation, and myomectomy rates, a short recovery time with rapid return
are frequently performed in ambulatory surgery to work, and shorter hospital stays and thus
settings,1 with many benefits for the patients and reduced costs.2-4

Lone Dragnes Brix, MHSc, RN, PhD student at the dations had no influence on the study design, collection, anal-
Department of Anesthesiology, Horsens Regional Hospital, ysis or interpretation of data, or in writing the article or in the
Horsens, Denmark; Theis Muncholm Thillemann, MD, PhD, decision to submit it for publication.
Speciality registrar at the Department of Orthopedic Surgery, Conflicts of interest: None to report.
Aarhus University Hospital, Aarhus, Denmark; and Lone Address correspondence to Lone Dragnes Brix, Department
Nikolajsen, MD, DMSc, Research consultant at the Danish of Anesthesiology, Horsens Regional Hospital, Sundvej 30,
Pain Research Centre and Department of Anesthesiology, Horsens 8700, Denmark; e-mail address: lonebrix@rm.dk.
Aarhus University Hospital, Aarhus, Denmark. Ó 2016 by American Society of PeriAnesthesia Nurses
L.D.B. has received funding from the Family Hede Nielsen 1089-9472/$36.00
Foundation, the Gurli and Hans Engell Friis Foundation, http://dx.doi.org/10.1016/j.jopan.2015.09.009
the Aase and Ejnar Danielsens Foundation (10-001290),
and the Health Research Fund of Central Denmark. The foun-

Journal of PeriAnesthesia Nursing, Vol -, No - (-), 2016: pp 1-8 1
2 BRIX, THILLEMANN, AND NIKOLAJSEN

Hysteroscopy is painful, and various methods for For patients in group LA 1 S, anesthesia was
anesthesia have been described, including paracer- induced with propofol 2.5 mg/kg/hour and remi-
vical local anesthetic injections combined with fentanil 12.5 mcg/kg/hour. When the patient
sedation and general anesthesia.5-7 Effective was able to sense the effect of remifentanil, infu-
postoperative pain management is essential for sion rates were reduced and maintained by a
early discharge after ambulatory surgery. continuous infusion of propofol 1.33 mg/kg/
However, there is no consensus on the choice of hour and remifentanil 3 to 5 mcg/kg/hour. The
anesthesia. We therefore decided to compare infusion rates were adjusted equal to a patient
two commonly used anesthetic methods: sedation level at 3 to 4 on the Ramsay Sedation
paracervical local anesthesia combined with Scale (1 to 6). Using a .4 3 35 long needle, the
sedation (LA 1 S) and general anesthesia (GA). gynecologist applied 3.60- to 7.2-mL Citanest
Dental Octapressin (30 mg/mL 1 .54 mcg/mL
Materials and Methods prilocaine hydrochloride 1 felypressin; Pierrel
S.p.A., Capua, Italy) for paracervical local anes-
Patients thesia. Bolus doses of 50- to 100-mcg fentanyl
IV were used for treatment of intraoperative
The study was approved by the Central Denmark Re-
pain $3 on a Numerical Rating Scale (NRS;
gion Committees on Health Research Ethics
0 5 no pain and 10 5 worst pain possible).
(M-20110023), the Danish Health and Medicines
Authority (EudraCT number: 2010-023843-13),
For patients in group GA, anesthesia was induced
and the Danish Data Protection Agency. It was
with propofol 10 mg/kg/hour and remifentanil
registered at Clinicaltrial.gov (NCT01412632),
50 mcg/kg/hour; and when the patient was able
conducted in accordance with guidelines for Good
to sense the effect of remifentanil, a bolus dose
Clinical Practice (GCP), and monitored by the
of propofol 2 to 2.5 mg/kg was given. Hereafter,
GCP-unit at Aarhus University Hospital, Denmark.
the infusions were maintained by continuous infu-
sion of propofol 5 mg/kg/hour and remifentanil
After obtaining written informed consent, 153
2.5 mcg/kg/hour. The infusion rates were adjusted
patients were enrolled at the Day Surgery Unit at
equal to a patient sedation level at 6 on Ramsay
Horsens Regional Hospital, Denmark. Inclusion
Sedation Scale, 1 to 6. The airway was managed
criteria were planned operative hysteroscopy
with a laryngeal mask, and the lungs were venti-
procedure and age $18 years. Exclusion criteria
lated with 40% to 45% oxygen in air. Bolus doses
were pregnancy, fibroids .3 3 3 cm, body mass
of 50- to 100-mcg fentanyl IV were used for
index $40, psychiatric illness, inability to commu-
intraoperative pain, guided by increases in pulse
nicate in Danish, or known allergy to remifentanil,
and blood pressure and/or motor unrest during
propofol, or prilocaine.
anesthesia. In both groups, remifentanil and pro-
pofol infusions were discontinued 4 to 5 minutes
Randomization and Intervention before the end of surgery. Surgery was performed
by experienced gynecologists specialized in hys-
On the day of surgery, patients were randomized to teroscopic procedures and trained in paracervical
receive LA 1 S or GA by using sequentially local anesthesia.
numbered opaque-sealed envelopes. Randomiza-
tion was performed in block sizes of 20 patients, The same team of anesthesia providers and nurses
each with a ratio of 1:1 according to a computer- in the PACU took care of and treated the two
generated sequence. Both groups received dexa- groups according to standard protocols regarding
methasone 8 mg, paracetamol 1 g, and ibuprofen pain and PONV/postdischarge nausea and vomit-
600 mg orally as premedication before anesthesia. ing (PDNV).
If ibuprofen was contraindicated, it was replaced
by tramadol 50 mg orally. According to the Care and Treatment
standard protocol of postoperative nausea and
vomiting (PONV) prophylaxis, intraoperative In the PACU, intensity of pain (NRS, 0-10) and
4-mg intravenous (IV) ondansetron was used for nausea and vomiting (yes/no) were assessed at
patients with a high risk of PONV. arrival, and every 15 minutes, or with shorter
LOCAL ANESTHESIA COMBINED WITH SEDATION 3

intervals if any changes in the intensity of pain, Normality of data was tested without pointing to-
nausea, and/or vomiting occurred. Pain in the ward normal distribution. Therefore, nonpara-
PACU (NRS $ 3) was treated with bolus doses of metric testing was performed. Fisher exact test
25- to 50-mcg fentanyl IV with a maximum of was used for dichotomous covariates and the
300 mcg in the PACU. PONV in the PACU was Wilcoxon rank sum test for continuous covariates.
treated with 4-mg ondansetron IV, .625-mg dehy- Results are presented as frequencies or medians
drobenzperidol (droperidol) IV, and 10-mg meto- with range as appropriate. All P values are two
clopramide IV. Shivering in the PACU was treated sided, and those below .05 were considered signif-
with 20-mg pethidine IV. icant. Intention-to-treat analysis was performed.
EpiData, version 3.1 (Epidata Association, Odense,
Oral postoperative analgesic treatment was started Denmark), was used for double entry, and statisti-
before discharge from the PACU and consisted of cal analysis was performed with Stata software,
paracetamol (acetaminophen) 1 g orally every version 12.0 (StataCorp, TX).
6 hours up to 4 g daily and ibuprofen 600 mg orally,
up to 1,800 mg daily as rescue medication. If Results
ibuprofen was contraindicated, it was replaced
by oral tramadol 50 mg, up to 200 mg daily. A total of 364 women were assessed for eligibility
Patients were discharged directly from the PACU from November 2011 through June 2013; 22 were
and were provided these oral analgesics to use at excluded based on the exclusion criteria, 157
home. declined to participate (72 preferred LA 1 S, 50
preferred GA, 35 declined to participate for other
Data Collection and Outcome Measures reasons), and 32 were excluded because of logistic
reasons. One hundred fifty-three patients were
Baseline patient characteristics were collected randomized: 76 to the LA 1 S group and 77 to
from medical records. The primary outcome mea- the GA group (Figure 1). Baseline characteristics
sure was the worst pain intensity score in the including age, height, weight, body mass index,
PACU rated by the patients on the NRS (0-10) smoking status, surgical procedure, and duration
from arrival to the PACU until discharge. Second- of surgery were comparable between groups
ary outcome measures were use of intraoperative (Table 1).
fentanyl (no/yes), use of analgesic and antiemetic
medication in the PACU (no/yes), PONV (no/yes), There was no significant difference in worst pain
time until discharge from PACU (minutes), recalled intensity between group LA 1 S and group GA in
worst pain after discharge (NRS, 0-10), and the PACU (1.5 [0 to 8] vs 2.2 [0 to 9], P 5 .13, me-
recalled PDNV (no/yes). dian [Range]; Figure 2A). Intraoperatively, fewer
patients in the LA 1 S group received treatment
Before discharge, all patients were supplied with a with fentanyl IV (14 vs 62; P , .01; Table 2). In
questionnaire including questions regarding pain the PACU, there was no statistical difference be-
and side effects during the 2 weeks after surgery. tween group LA 1 S and group GA in the number
The questionnaire was a log for the investigator of patients treated with fentanyl IV (6 vs 13;
to document the patients’ verbal response when P 5 .13; Table 2). There was no significant differ-
they were contacted by phone 2 weeks after sur- ence in the number of patients who experienced
gery and asked to recall the worst pain after nausea in the PACU between group LA 1 S and
discharge from PACU and PDNV. group GA (4 vs 4; P 5 .90) or the number of
patients who received ondansetron IV (4 vs 4;
Statistical Analysis P 5 .90). In group LA 1 S, fewer patients received
PONV prophylaxis (5 vs 31; P , .01). None of the
The sample size calculation was based on an patients vomited during hospitalization (Table 3).
expected difference of 1 on the NRS with 80% po- Time in the PACU (63 [15 to 210] vs 81 [15 to
wer (a 5 .05, b 5 .2). A sample size of 47 per 200] minutes, P , .01, median [range]) and time
group was required, ie, a total of 94 patients. Based from start of surgery until discharge (99 [40 to
on an expected response rate of 65% after 2 weeks, 305] vs 121 [60 to 265] minutes, P , .01) were
it was decided to include at least 150 patients. significantly shorter in group LA 1 S group.
4 BRIX, THILLEMANN, AND NIKOLAJSEN

Analyzed Analyzed

Figure 1. CONSORT study flow diagram. BMI, body mass index; CONSORT, CONsolidated Standards of Reporting
Trials; GA, general anesthesia; LA 1 S, local anesthesia combined with sedation.

Two weeks after discharge from PACU, there was Discussion
no significant difference in recalled worst pain
intensity between group LA 1 S and group GA In this study of ambulatory operative hysteroscopy
(1.5 [0 to 8] vs 1.5 [0 to 8], P 5 .40, median procedures, LA 1 S did not significantly decrease
[Range]; Figure 2B). A total of 12 patients, five patients’ worst pain intensity in the PACU. Howev-
in the LA 1 S and seven in group GA, recalled er, the number of patients treated with IV fentanyl
having suffered from nausea (P 5 .46) and four intraoperatively and time from start of surgery un-
patients, all belonging to group LA 1 S, recalled til discharge from PACU were significantly reduced
having vomited after discharge (P 5 .05; in group LA 1 S. More patients in group LA 1 S
Table 3). vomited after discharge.
LOCAL ANESTHESIA COMBINED WITH SEDATION 5

Table 1. Baseline Demographic and Clinical Characteristics of Study Participants
Baseline Characteristics LA 1 S (n 5 69) GA (n 5 75) P value
Age (y) 47 45 .05
[29 to 78] [27 to 67]
Height (m) 1.67 1.68 .28
[153 to 185] [150 to 187]
Weight (kg) 71 kg 73 kg .97
[48 to 100] [50 to 111]
BMI (kg/m2) 25.2 24.6 .93
[18.6 to 37.2] [19.2 to 38.1]
Smoking status (no/yes/missing) 53/16/0 63/10/2 .14
Surgical procedure
Endometrial ablation 33 46 .06
Myomectomy 2 2
Polymectomy 22 19
Combined operations 12 8
Duration of surgery (min) 35 37.5 .07
[15 to 95] [20 to 80]
BMI, body mass index; GA, general anesthesia; LA 1 S, local anesthesia combined with sedation; PACU, postanes-
thetic care unit.
Patient characteristics in the interventions group (LA 1 S) and in the control group (GA).
Median [range].

Some methodological issues must be considered. be judged as having received either LA 1 S or GA
First is the lack of blinding. Attempts to blind the just by observation on arrival in the PACU. Finally,
study would have failed for several reasons. In most patients knew how they were anesthetized.
the operating room, blinding was obviously not
possible and in the PACU, the organization of the Second, the lack of difference can be attributed to
unit, with the nursing staff working in both the low pain intensity score in both groups. In the pre-
operating room and in the PACU made blinding sent study, both groups received propofol, which
impossible. In addition, almost every patient could is associated with significantly less postoperative
40

40
30

30
Number of patients

Number of patients
20

20
10

10
0

0

0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10
NRS, 0-10 NRS, 0-10

LA+S GA LA+S GA

Figure 2. A1B: Maximum pain intensity in the PACU (A) and after discharge (B) The maximum pain intensity
scores in the PACU (A) and after discharge (B) after ambulatory operative hysteroscopy. GA, general anesthesia;
LA 1 S, local anesthesia combined with sedation; NRS 0-10, Numerical Rating Scale from 0 5 no pain to
10 5 worst pain possible; PACU 5 postanesthesia care unit.
6 BRIX, THILLEMANN, AND NIKOLAJSEN

Table 2. Number of Patients Who Received Fentanyl Intraoperatively and in the PACU
Analgesic medication LA 1 S (n 5 69) GA (n 5 75) P Value
Fentanyl IV intraoperatively (no/yes) 55/14 13/62 , .01
Fentanyl IV in PACU (no/yes) 63/6 62/13 .13
GA, general anesthesia; IV, intravenous; LA 1 S, local anesthesia with sedation; PACU, postanesthetic care unit.
Bold indicates significant level was set at P , 0.05.

pain. Combined with a short duration of surgery, Compared with previous studies, we found that
this may have contributed to the low pain inten- patients in both groups had a lower incidence of
sity.8-12 A presumably large proportion of nausea in the PACU12,17 and after discharge.18
patients declined to participate because they had The lower incidence of nausea in the PACU
strong preferences regarding either general is, however, similar to recent findings showing
anesthesia or local anesthesia combined with that patients who were anesthetized with
sedation. Being given a choice of anesthetic remifentanil-propofol had less postoperative nausea
induction method can reduce anxiety,13 and the than patients anesthetized with remifentanil only.19
enrolled patients may therefore have a higher level
of anxiety and pain intensity than all patients A previous study found as many as 36% of patients
eligible for the study. The most anxious patients who experienced PDNV did not experience
may have chosen not to participate in the study. nausea and vomiting before discharge from hospi-
However, the low pain score is in accordance tal.17 In the present study, only patients in group
with other studies which found that gynecological LA 1 S vomited after discharge. This could be
patients have fewer postoperative symptoms, attributed to the lower number of patients in the
including postoperative pain, than orthopaedic LA 1 S group who received PONV prophylaxis
and general surgery patients.14-16 because of a lower risk of PONV. PDNV is consid-
ered to be a growing concern, and suitable strate-
Third, patients were interviewed by telephone gies should be planned for patients after discharge
2 weeks after surgery regarding the worst pain in- because patients often do not have access to
tensity after discharge. The pain intensity score antiemetic treatment after discharge.11,17
after discharge may be influenced by recall bias, Preventing PDNV is important because it may
due to the fact that patients were asked to recall delay the resumption of normal activities,
the worst pain after discharge from PACU on the readiness for work, and have a negative impact
NRS. on the patient’s experience and recovery.9,12

Table 3. Number of Patients Who Experienced PONV, PDNV, and Received Antiemetic Treatment
PONV LA 1 S (n 5 69) GA (n 5 75) P Value
Ondansetron (no/yes) intraop 64/5 44/31 , .01
Nausea (no/yes) 66/4 71/4 .90
Vomiting (no/yes) 69/0 75/0 —
Ondansetron (no/yes) PACU 65/4 71/4 .90
DHB (no/yes) 70/0 74/1 .52
PDNV LA 1 S (n 5 62) GA (n 5 69)
Nausea (no/yes) 57/5 62/7 .46
Vomiting (no/yes) 58/4 69/0 , .05
GA, general anesthesia; DHB, dehydrobenzperidol; LA 1 S, local anesthesia combined with sedation; PACU, postanes-
thesia care unit; PDNV, postdischarge nausea and vomiting; PONV, postoperative nausea and vomiting.
The number of patients who experienced PONV, PDNV, and received antiemetic treatment.
LOCAL ANESTHESIA COMBINED WITH SEDATION 7

As in the present study, another Danish study patients more closely during the follow-up period
found a significantly reduced time from surgery and preoperative pain, psychological factors, and
until discharge in patients who were anesthetized time before resumptions of normal activities after
with paracervical local anesthesia with sedation discharge need to be taken into account.
compared to patients who were anesthetized
with general anesthesia for ambulatory operative
hysteroscopy.19
Conclusion

In summary, no statistically significant difference
Although some methodological issues exist, as between the groups was evident for postoperative
described, the present study is distinguished by pain. However, local anesthesia combined with
successful randomization and adherence to the sedation with remifentanil and propofol could be
principles of GCP. No serious side effects were recommended as the first choice of anesthetic
observed. We find external validity to be strength- technique for patients undergoing ambulatory
ened by the fact that the study was carried out in a operative hysteroscopy because of the decreased
daily clinical setting with the anesthesia protocols number of patients treated with IV fentanyl intrao-
and discharge criteria reflecting daily practice. peratively and the reduced time from start of
Overall, the findings of the study can be general- surgery until discharge from PACU. Increased
ized to women having operative hysteroscopy pro- emphasis on PDNV prophylaxis must be consid-
cedures in an ambulatory surgery setting. ered for these patients.
Recommendations for future studies include hav-
ing a smaller team of specially trained providers Acknowledgments
who are responsible for data collection and a
The authors would like to thank the patients who participated
data collection instrument using a series of in this study, the staff at the Department of Gynecology and
possible responses instead of dichotomous Obstetrics and the Department of Anesthesiology at Horsens
answers. Further studies should also monitor Regional Hospital.

References
1. Nathani F, Clark TJ. Uterine polypectomy in the manage- systematic review and meta-analysis. Anaesthesia. 2014;69:
ment of abnormal uterine bleeding: A systematic review. 1138-1150.
J Minim Invasive Gynecol. 2006;13:260-268. 10. Li M, Mei W, Wang P, et al. Propofol reduces early post-
2. Fothergill RE. Endometrial ablation in the office setting. operative pain after gynecological laparoscopy. Acta Anaesthe-
Obstet Gynecol Clin North Am. 2008;35:317-330. x. siol Scand. 2012;56:368-375.
3. Kremer C, Duffy S, Moroney M. Patient satisfaction with 11. Gan TJ. Consensus guidelines for the management of
outpatient hysteroscopy versus day case hysteroscopy: Rando- postoperative nausea and vomiting. Anesth Analg. 2014;118:
mised controlled trial. BMJ. 2000;320:279-282. 85-113.
4. O’Flynn H, Murphy LL, Ahmad G, Watson AJ. Pain re- 12. Awad IT, Chung F. Factors affecting recovery and
lief in outpatient hysteroscopy: A survey of current UK discharge following ambulatory surgery. Can J Anaesth. 2006;
clinical practice. Eur J Obstet Gynecol Reprod Biol. 53:858-872.
2011;154:9-15. 13. Mitchell M. Influence of gender and anaesthesia type on
5. Cooper NA, Khan KS, Clark TJ. Local anaesthesia for pain day surgery anxiety. J Adv Nurs. 2012;68:1014-1025.
control during outpatient hysteroscopy: Systematic review and 14. Brattwall M. Patient assessed health profile: A six-month
meta-analysis. BMJ. 2010;340:c1130. quality of life questionnaire survey after day surgery. Scand
6. Goulson DT. Anesthesia for outpatient gynecologic sur- J Public Health. 2010;38:574-579.
gery. Curr Opin Anaesthesiol. 2007;20:195-200. 15. Berg K, Kjellgren K, Unosson M, Arestedt K. Postop-
7. Wallage S, Cooper KG, Graham WJ, Parkin DE. A rando- erative recovery and its association with health-related
mised trial comparing local versus general anaesthesia for quality of life among day surgery patients. BMC Nurs.
microwave endometrial ablation. BJOG. 2003;110:799-807. 2012;11:24.
8. Eger EI, White PF, Bogetz MS. Clinical and economic 16. Chung F, Ritchie E, Su J. Postoperative pain in ambulatory
factors important to anaesthetic choice for day-case surgery. surgery. Anesth Analg. 1997;85:808-816.
Pharmacoeconomics. 2000;17:245-262. 17. Wu CL, Berenholtz SM, Pronovost PJ, Fleisher LA.
9. Kumar G, Stendall C, Mistry R, Gurusamy K, Walker D. Systematic review and analysis of postdischarge symptoms
A comparison of total intravenous anaesthesia using propofol after outpatient surgery. Anesthesiology. 2002;96:
with sevoflurane or desflurane in ambulatory surgery: 994-1003.
8 BRIX, THILLEMANN, AND NIKOLAJSEN

18. Gupta A, Wu CL, Elkassabany N, Krug CE, Parker SD, 19. Majholm B, Bartholdy J, Clausen HV, Virkus RA,
Fleisher LA. Does the routine prophylactic use of antiemetics Engbaek J, Moller AM. Comparison between local anaesthesia
affect the incidence of postdischarge nausea and vomiting with remifentanil and total intravenous anaesthesia for opera-
following ambulatory surgery?: A systematic review of random- tive hysteroscopic procedures in day surgery. Br J Anaesth.
ized controlled trials. Anesthesiology. 2003;99:488-495. 2012;108:245-253.