Mcr

ID

a

ARRA

KPPMII

1

oeet

tInci

LT

1h

Pharmacological Research 66 (2012) 185– 191

Contents lists available at SciVerse ScienceDirect

Pharmacological Research

jo ur n al hom epage: www.elsev ier .com/ locate /yphrs

orphine at “sub-analgesic” background infusion rate plus low-dose PCA bolusontrol pain better and is as safe as twice a bolus-only PCA regimen: Aandomized, double blind study

an White, Ronen Ghinea, Shmuel Avital, Shoshana Chazan, Oleg Dolkart, Avi A. Weinbroum ∗

epartment Surgery A, and Post-Anesthesia Care Unit, Tel Aviv Sourasky Medical Center, Tel Aviv; the Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel

r t i c l e i n f o

rticle history:eceived 19 March 2012eceived in revised form 29 March 2012ccepted 29 March 2012

eywords:ainostoperativeorphine

V-PCAnfusion

a b s t r a c t

Morphine for postoperative pain control is commonly titrated via intravenous patient-controlled anal-gesia (IV-PCA). An IV morphine background infusion is rarely used. We investigated whether analgesia iseffectively attained and morphine consumption is reduced if PCA titration is coadjuvated by a continuousinfusion protocol. Following colorectal cancer surgery, consenting patients were randomized to receivea minimal (“sub-analgesic”) dose of morphine 0.01 mg/kg/h background infusion plus a 0.01 mg/kg bolus(BI), or a 1.5 mg bolus-only morphine (B0) (bolus ratio ∼1:2). Bolus lockout time was 7 min in eithercase. All patients received 0.1 mg/kg morphine before protocol initiation, and diclofenac 75 mg intra-muscularly b.i.d. during the study period, lasting 48 h. Eighty-six patients (51 males, age 26–95 years)participated in the study. The total mean morphine consumption during the 48 h was 25% lower in the BIthan in the B0 group (P < 0.05). Although the former applied the PCA device for boluses 19% less than thelatter (P < 0.05), their pain score was lower (P < 0.05) most of the time, and they reported greater satis-faction (P < 0.05) on a 10-scale numerical rating score. Pre- and postoperative vital signs were similar forboth groups. No patient depicted hypoxemia or lapsed into deep sedation. Four BI and three B0 patients

required treatment for postoperative nausea and vomiting. One BI patient had transient pruritus and oneB0 69-year individual became disoriented 24 h into treatment; either event subsided soon after stoppingtheir respective regimen without the need for treatment. The main conclusions of the results are thatvery-low-dose background morphine infusion combined with small-dose PCA boluses may provide bet-ter pain relief, lower morphine consumption, and minimal complication rate as a 1.5 mg PCA bolus-onlyprotocol.

© 2012 Elsevier Ltd. All rights reserved.

. Introduction

Postoperative pain management is still a challenge to cliniciansf various medical fields. Morphine is the most widely used postop-

rative analgesic. Nevertheless, concerns still exist regarding sideffects, although these and its efficacy are only moderate comparedo other perioperatively used opioids [1]. Its reputation probably

Abbreviations: ASA, American Society of Anesthesiologists; B0, bolus only; Cemax,he maximum concentration in the effect compartment; ICU, intensive care unit;M, intramuscularly; IV, intravenous; MAC, minimal alveolar concentration; NRS,umerical rating scale; OIH, opioid-induced hyperalgesia; PACU, post-anesthesiaare unit; PCA, patient-controlled analgesia; PONV, postoperative nausea and vom-ting; SpO2, pulse-derived oxygen saturation; VAS, visual analog scale.∗ Corresponding author at: Post-Anesthesia Care Unit and Pre-Clinical Research

aboratory, Tel Aviv Medical Center, 6 Weizman St., Tel Aviv 64239, Israel.el.: +972 3 6973237; fax: +972 3 6925749.

E-mail address: draviw@tasmc.heatlh.gov.il (A.A. Weinbroum).

043-6618/$ – see front matter © 2012 Elsevier Ltd. All rights reserved.ttp://dx.doi.org/10.1016/j.phrs.2012.03.016

lies in the fact that it is less potent and therefore risky than syntheticopioids, such as fentanyl or alfentanyl, mainly for naïve or elderlyindividuals, and where close monitoring of vital signs (e.g., bloodpressure, respiratory rate, and oxygenation) are seldom availableas on the ward.

Patient-controlled analgesia (PCA) is the most common modeof intravenous (IV) titration of opioids for controlling moderate-to-severe postoperative pain. Although the administration of smallIV boluses (∼1.0–2.5 mg) allows for rapid and safe titration of thedose needed for adequate pain relief, adverse effects have stillbeen reported, among them sedation, respiratory depression andhypoxia. Lockout time is implemented to prevent these untowardsequelae, and its duration is fixed between 3 and 10 min, dependingon the patient’s characteristics, the type of opioid and its pharma-

cology, and the amount of each bolus.

The IV-PCA bolus dose of morphine (and other opioids) dependson many factors, such as pain intensity, previous use of opioids,and co-morbidities. In patients under close observation, as in the

1 ical R

sumvbrlslrecruafbs[

trtbatsvp

2

sofcaedt[ca

ssplaunsPrmitaqticri

86 I. White et al. / Pharmacolog

etting of a post-anesthesia care unit (PACU) or intensive carenit (ICU), a continuous infusion of the opioid has been advocatedainly in children [1–3], under the presumption that it would pro-

ide constant anti-nociception, while minimizing adverse eventsy curtailing fluctuations in blood concentration [4]. It has beeneported that a standard dose of morphine infusion may accumu-ate in hypovolemic or elderly patients, as well as in individualsuffering from kidney or liver dysfunction if administration is pro-onged [5]. Another disadvantage is that infusion is less readilye-adjustable in cases where pain fluctuates during the postop-rative course. One example is large bowel surgery, after whichonsiderable pain can be expected for several days during earlyecovery. We have shown that a combination of an infusion topped-p with boluses could profit the latter cases, by providing rapid,dequate and safe pharmacological response to intervallic needsor postoperative pain [6]. Sucato et al. have reported the com-ination of a basal infusion of morphine in their PCA protocol inpine-operated patients, which was kept constant at 0.015 mg/kg/h7].

Due to the inconsistency of opinions regarding the above pro-ocols, the acknowledgement of the utility of background plus PCAegimen, while no large randomized studies have been undertakeno compare such protocols, we performed a prospective, double-lind, randomized study aiming to evaluate the antinociceptivedded value and the occurrence of side effects of minimal con-inuous background infusion of morphine when combined withelf-administered IV-PCA titrated low-dose boluses of morphine,ersus twice the bolus dose in a bolus-only titration mode, inatients who underwent colorectal surgery.

. Patients and methods

After obtaining the local institutional review board (IRB) permis-ion to conduct this study, 100 patients were invited to participate,f whom 90 agreed and signed the IRB-approved informed consentorm. There was no age limit, given that the elderly are an espe-ially suitable target population of this study for their pathology,nd both complications and optimal pain control. During the preop-rative interview, all the patients were taught how to discern whatefined an unacceptable level of pain, and instructed to request ini-iation of the PCA setup if and when their pain reached that level8]. Each patient was shown how to use a standard 10-cm numeri-al rating scale (NRS) to enable postoperative pain and satisfactionssessments.

All the patients underwent the same surgical procedure undertandardized general anesthesia, and by the same teams. Exclu-ion criteria included American Society of Anesthesiologists (ASA)hysical class >3, non-colorectal abdominal surgery, planned pro-

onged mechanical ventilation or an extended stay in an ICU forny reason, history of drug abuse or chronic pain, patients sched-led for minimally invasive surgery (e.g., laparoscopic surgery) oreeding urgent intervention, or those known to have hypersen-itivity to any drug administered during the 48-h study period.atients with severe liver, cardiac, renal or pulmonary disease,ecent (<6 mo) cerebrovascular accident or cardiac event, orental incapacity, were also excluded. Subjects were dropped

f they required >4 h of unplanned postoperative assisted ven-ilation, underwent re-operation during the study period, hadd-hoc need for ICU transfer, were incoherent or lacked ade-uate comprehension of their surroundings after surgery, andhose for whom there were intra-operatively decided changes

n surgical plan. Other cause for a later disqualification wereore temperature <35.0 ◦C upon arrival to the PACU, pain notelieved by the applied regimen, and the presence of continu-ng (>15 min) blood de-saturation (<92% on 40% oxygen mask).

esearch 66 (2012) 185– 191

The data of all those subjects were not included in the final analy-ses.

All patients were given IV midazolam up to 0.05 mg/kg, fentanyl1.5 �g/kg, propofol 1–1.5 mg/kg and 0.7 mg/kg of rocuronium forthe induction of anesthesia. Maintenance consisted of isoflurane-enriched nitrous oxide-in-oxygen 66/33%, aiming at end-tidalminimal alveolar concentration (MAC) of 0.8–1.2. A muscle relaxantwas added as deemed necessary, and the fentanyl dosage was fur-ther adjusted to hemodynamic drifts and signs of pain. At the end ofsurgery, relaxation was reversed and the patients were transferredto the PACU.

2.1. Randomization

The quality of the study was assessed using the following Jadadcriteria: random allocation of treatments with a clear descriptionof the randomization procedure, blinding of the patient for theassigned treatment, blinding of the outcome assessor, and descrip-tion of dropouts and missing values [9]. Following the surgicalprocedure, the PCA device was connected to the patient and startedby the attending anesthesiologist. The syringe and the device wereprepared and programmed based on the randomization list in theinstitutional pharmacy and by an individual uninvolved in thestudy, respectively. In case of complication, only one researcher (A.A. W.) had access to the computerized list of study group assign-ment. In addition, the data of patients who were withdrawn or whodropped out were incorporated in the ‘intention-to-treat’ analysesof the baseline data, and their outcomes were omitted from furtherassessment. Any patient could quit the study for any reason, andhis/her data were not used for analyses.

2.2. Drug protocols and study goals

The first series of vital sign measurements and confirmation ofthe patient’s coherence were obtained upon arrival to the PACUfrom the operating room. Following our standard protocol for post-operative care, patients in both study groups received a titratedloading dose of IV morphine 0.1 mg/kg during a 15-min period.This allowed all patients to start PCA use while under a similarnociception–antinociceptive equilibrium. The IV-PCA device wasconnected to the IV line of the patient as soon as he/she reportedof pain that requested analgesia. The device was programmed todeliver either a 1.5 mg/bolus morphine as deemed necessary to thepatient, to be comfortable with pain at deep breath, with a lockouttime of 7 min (the B0 group), or a 0.01 mg/kg bolus at the samecriteria, in combination with a continuous background infusionof 0.01 mg/kg/h of morphine (the BI group) [1]. The B0 protocolrepresents our institution’s long-standing protocol, which is basedon pharmacological application of a mean of 0.02 mg/kg morphinein individuals weighing about 70–75 kg [1,3,10–14]. The infusionhourly dose was based on earlier reports of infusion of morphinein the immediate postoperative period [1,15]. In order to evalu-ate the added value of the infusion, the BI protocol applied only50% the B0 bolus dose, so that antinociceptive pharmacologicaleffect was evident, and infusion-dependent safety could be main-tained, while effective titration was available for sporadic painintensifications.

Diclofenac 75 mg IM was the rescue drug available to eachpatient, and it was given at fixed times twice daily, starting at1 h after starting the PCA regimen. This enabled the evaluation ofthe analgesic effect of morphine in either group, specifically thecontribution of the background infusion.

All the patients were transferred to the surgical ward 3 hafter starting the regimen, assuring for enough time to reach lev-els of pain high enough to be eligible to start using the PCAdevice, and to allow close monitoring of its safe implementation

ical Research 66 (2012) 185– 191 187

dwmatrsp1s{mitp

ubpsftfinat

2

fpfcarompbcoattvmcttad

3

ioswPprs

Excluded ( n=0)

Not meetin g inclusion criteria (n=5)

Refused to participate (n=5)

Other reasons (n=0)

Assessed for Eligibilit y (n=100)

Consenting, Rando mized (n=90)

All ocate d t o B0 (n=41 )

Received all ocated intervention (n=41 )

Allocated to BI (n=45)

Received all ocated intervention (n=45 )

Lost to fol low-up (n=0)

Discontinued intervention (n=1)

Conversion (n=0)

Lost to fol low-up (n=0)

Discontinued intervention (n=1)

Conversion (n=0)

Analyzed (n=41 )

Intended-to-treat analysis (n=1)

Dropped out (n=4)

Analyzed (n=45 )

Intended-to-treat analysis (n=1)

Dropped out (n=0)

Analysis

Follow-up

Allocation

Enrolment

I. White et al. / Pharmacolog

uring the initial phase. The device was available for 48 h, afterhich the antinociceptive regimen followed the surgical depart-ent’s protocol. Pain intensity during breathing was evaluated

nd recorded using the NRS twice daily and at 6 h postopera-ively. Any side effects such as nausea, vomiting, pruritus, urinaryetention or dizziness, were recorded and treated. Severe nau-ea or vomiting was treated with 10 mg metoclopramide; severeruritus caused stopping the PCA protocol and was treated with0 mg IV promethazine if still required. If respiratory depres-ion occurred (<6 breaths/min or pulse-derived oxygen saturationSpO2} <92% under 40% oxygen mask), PCA was stopped, and inter-ittent doses of 0.04 mg of naloxone were administered at 2-min

ntervals until the condition reversed. Patient satisfaction usinghe NRS (1–10 score) was also recorded at the end of the studyeriod.

The two primary end-points of this study were the 48-h effectivesage of the device and the total amounts of morphine consumedy each group during that period. The secondary goals were (1)ain scores during the study period, (2) the final (i.e., the 48-h)atisfaction rate, (3) parameters referring to the speed of recoveryrom surgery that can be influenced by postoperative antinocicep-ive drugs, i.e.: first time of leaving bed, first passage of flatus,rst passage of stools, removal of urinary catheter, and removal ofasogastric tube, (4) incidences of morphine- and NSAID-relateddverse events, and (5) number of patients wanting to drop out ofhe study.

.3. Statistics

The statistical analyses were performed using the SPSS Releaseor Windows, Version 14.01 (Chicago, IL, 2008). The pre-studyower table (using a 2-sample t-test) consisting of delta (the dif-erence in morphine use during the first 6 postoperative hoursalculated in a pilot study) 1.0 ± 0.8, and alpha (type I error prob-bility associated with this test of the null hypothesis) ≤0.05,esulted in the need for a minimum of 40 patients per group inrder to be able to reject the null hypothesis that the populationeans of the experimental and control groups are equal, with a

robability (power) of 0.8. The demographic data (age, weight) andackground characteristics (amounts of intra-operative fentanylonsumption, the duration of surgery) were compared using thene-way ANOVA. Gender distributions, the incidence of side effects,nd the rate of drop-out or complications were analyzed usinghe Fisher exact test. All repeated values, i.e., the hourly consump-ion of morphine, the rate of device applications, pain scores, andital signs, were analyzed by the one-way ANOVA with repeatedeasures. The t-test was employed to analyze the total morphine

onsumption during the study period as well as the final satisfac-ion rate of the patients. The ANOVA test was always followed byhe post hoc Tukey’s Honest Significant Difference method. Valuesre expressed as mean ± SD, or absolute numbers, with significanceefined as P ≤ 0.05.

. Results

Fig. 1 depicts the CONSORT flowchart. Of the 90 consent-ng patients, 86 completed the study, with four having droppedut postoperatively. The demographic and background data wereimilar between the two groups (Table 1), and all the patientsere awake and communicated with the medical staff in the

ACU 15 min after their arrival. Hemodynamic and respiratoryarameters, including oxygenation, were all within the normalanges in both groups throughout the study period (data nothown).

Fig. 1. The CONSORT flowchart.

3.1. Primary end-points

The rates of PCA delivery were different between the groups.The effective rates of the device’s application (which actually deliv-ers the drug) declined during the first postoperative hours in bothgroups, after which they stabilized. Drug deliveries were morestable over time in the BI group (Fig. 2). The BI patients made sig-nificantly less use of the device than the B0 patients throughout the48-h study period (Table 2). Overall, the BI group applied the PCAdevice for boluses 19% less than the B0 group. The most evidentdrop in device use occurred between 24 and 48 h postoperatively,when the mean hourly usage per person in the BI group was 50% ofwhat it had been during the 0–24 h period, while B0 patients’ usagewas reduced by only 20% (Table 2). Finally, unlike the fluctuationin the rate of self-administration of morphine observable in the B0group, the BI patients activated the device at a constant rate formany hours (Fig. 2).

The hourly consumption of morphine by the BI patients wassignificantly less than that of the B0 patients, starting at hour 3of treatment (Fig. 2). This difference was maintained throughoutthe 48-h study period (Figs. 2 and 3). As depicted in Fig. 4, the

total amount of morphine used as boluses by the BI patients was16.8 ± 11.7 mg compared to 50.9 ± 26.2 mg used by the B0 patients(P = 0.034). Furthermore, despite the protocols’ given preamble thatthe mean PCA-delivered bolus dose in the BI group approximated

188 I. White et al. / Pharmacological Research 66 (2012) 185– 191

Table 1Demographic and background data (Mean ± SD or absolute values).

Variables Bolus only (45 patients) Bolus + infusion (45 patients) P value

Age (yr) 62 ± 13 65 ± 10 0.16Males/Females (n) 23/18 28/17 0.56Weight (kg) 76 ± 13 71 ± 11 0.078Duration of surgery (min) 216 ± 64 198 ± 68 0.27Intraoperative fentanyl consumption (�g/patient) 428 ± 204 397 ± 162 0.47

Table 2Postoperative parameters (Mean ± SD or absolute values).

Variables Bolus only (41 patients) Bolus + infusion (45 patients) P value

PACU cumulative pain score (NRS 0–10) 5.72 ± 1.27 3.41 ± 1.19 <0.0001PACU cumulative sedation score (NRS 1–10) 6.8 ± 1.55 5.38 ± 1.93 0.0470–24 hourly effective use of PCA device (n) 1.07 ± 1.24 0.92 ± 1.6 0.0224–48 hourly effective use of PCA device (n) 0.83 ± 1.1 0.52 ± 0.98 <0.000148-h cumulative pain scores (NRS 0–10) 4.07 ± 2.0 1.40 ± 2.0 <0.0001Postoperative global satisfaction score (1–10) 4.69 ± 2.09 6.95 ± 1.82 <0.0001First time out of bed (day) 1.65 ± 1.11 1.18 ± 0.39 0.02First bowel movement (flatus) (day) 3.52 ± 1.59 3.61 ± 1.37 0.8First stool passage (day) 4.55 ± 2.25 5.05 ± 2.18 0.35Time to urinary catheter removal (day) 4.2 ± 1.54 4.21 ± 2.67 0.98Time to nasogastric tube removal (day) 2.19 ± 1.74 2.08 ± 1.57 0.77Postoperative nausea/vomitinga (n) 3 4 0.79Surgical complications (>7<30 days) 10 11 1

P cal rat

oafpi

3

fiBsoc

r

Fsa

Eleven BI patients and ten B0 patients suffered from surgical

CA = patient-controlled analgesia; PACU = post-anesthesia care unit; NRS = numeria Requiring treatment.

ne-half of that of the B0 group (i.e., 0.7 vs. 1.5, mg considering mean 70-kg body weight), calculations of the weight-to-doseactors revealed a true 25% lower morphine consumption by BIatients compared to B0 patients, the infused amount of morphine

ncluded (Figs. 3 and 4).

.2. Secondary end-points

The periodic mean pain scores and the entire period cumulativegure, as well as the final satisfaction rates, were all better for theI group than for the B0 group (Fig. 5, Table 2). The 48-h mean paincore was similar between the two groups (Fig. 5). The proportionf pain NRS >5/10 detected on ward was 24% among the B0 patients

ompared to 7% in the BI group (P = 0.001).

The parameters characterizing the speed by which patientsecovered from surgery were all better for the BI group than for the

60 12 18 24 30 36 42 48

-1

0

1

2

3

4

5

a

Mo

rph

ine

ho

urly c

on

su

mp

tio

n

by b

olu

se

s (

mg

/h)

Time (hours)

Bolus+Infusion group

Bolus-only group

ig. 2. Hourly consumption of morphine by boluses (mean ± SD). Morphine con-umption directly reflects the hourly PCA application rates in either group.P < 0.0001.

ing score.

B0 group, with the time to first leave bed freely reaching a statisticalsignificance (Table 2).

Adverse effects in the 48 postoperative hours occurred rarely.One BI patient suffered from pruritus, and one B0 patient (69 yearsof age) was disoriented 24 h after being connected to the PCAdevice: both were disconnected from the device, and the eventssubsided several hours later without the need for treatment or theoccurrence of further sequelae. Three B0 and four BI patients werereported to have suffered from postoperative nausea and vomit-ing (PONV) that required one-time treatment. There were no otheradverse events in either group. No patient asked to leave the study;none developed deep sedation, bradypnea or de-saturation, and nopatient needed ICU admission.

complications that did not appear until day 7 after the proce-dure (Table 2), which, therefore, did not interfere with the currentstudy.

Fig. 3. Consumption of morphine at various time points during the study period(mean ± SD). aP < 0.0001.

I. White et al. / Pharmacological R

Fig. 4. Total morphine consumption (mean ± SD). aP = 0.034; bP < 0.0001.

Fm

4

cdbftthreWtePa

rietph

ig. 5. Periodic pain scores (mean ± SD). The “3” (three) h values indicate the PACUean cumulative pain score. aP = 0.01.

. Discussion

The results of this randomized, double blind, clinical trial indi-ate that the administration of one-half the commonly used bolusose of morphine, i.e., 0.01 mg/kg, in combination with a minimalackground infusion dose (0.01 mg/kg/h, or 0.7 mg/h as calculatedor a ∼70-kg individual) was more effective in reducing pain thanhe two-fold higher bolus dose without background infusion. Fur-hermore, the low-dose study protocol was associated with lowerourly and total morphine consumption, a higher final satisfactionate, and a similarly low incidence of morphine-related adversevents during the most painful period, i.e., 48 h following surgery.e propose that low background infusion dose of morphine,

hat would not have induced any clinical or pharmacologicalffect if administered solely, can enhance the antinociceptiveCA-morphine titration during the first two postoperative daysfter colorectal surgery.

Protocols for postoperative pain control have undergoneeassessments during recent years, focusing on the two majorssues, i.e., optimization of pain control and lessening adverse

vents [4,10–14]. PCA is a more efficacious and safer mode of sys-emic postoperative pain control on the ward than other parenteralrotocols (e.g., intramuscular administration). PCA protocols differ,owever, with regard to the bolus dose, the lockout time, and

esearch 66 (2012) 185– 191 189

the duration of the device’s availability to the patient. Theseparameters depend on a number of variables, such as the patient’sage, concomitant pathologies, type, and extent of surgery, andsurgeon’s preferences. Our group has long been using a protocolthat allows the administration of a dose of 1.5 mg (∼0.02 mg/kgin an average 70-kg patient) of morphine every 7 min [10–14,16].There are very few reports on the implementation of protocolsthat combine morphine titration on top of its background infusion.The use of a concurrent background infusion with PCA in adults isan area of debate in the literature (see below). The clinical valuesof such protocols have thus not been established, and the resultsof their use remain uncertain.

In order to study the pro-analgesic effects of the low-dose mor-phine infusion (plus titration), it was necessary to minimize fluc-tuating pharmacological interference that results from the flexibleadministration (i.e., by request) of a rescue drug. Accurate analysesof the analgesic effects of low-dose morphine would require theelimination of any rescue drug, which, however, is not ethicallypossible. Alternatively, their analgesic effects should be strictlycontrolled by scheduling their administration to the patient, thusmaintaining a constant pharmacological effect on pain. Further-more, NSAIDs by themselves, even in large doses, do not providesufficient pain relief after major surgery, although they are believedto offer variable opioid-sparing effects [17]. For this reason, all thepatients of both groups were given intramuscular diclofenac 75 mgb.i.d. as the rescue drug during the 48 postoperative hours.

The early “boosting dose” of morphine permitted the initi-ation of PCA delivery when all the patients were in a similarnociception–antinociception state. This was done although weassumed the amounts of intraoperative fentanyl would be simi-lar in both groups, and although the same surgical and anesthesiateams handled all cases. Also, a PCA regimen would not controlpain efficaciously unless an adequate pharmacological blood levelof the opioid has been built previously. According to Lötsch et al., themean dose for unacceptable pain relief, as for our pre-PCA boostingdose, is 0.17 ± 0.10 mg/kg morphine [16]. Therefore, we consideredthat titration to 0.1 mg/kg dose as being appropriate before startingPCA. Indeed, Chen et al. also administered a 4-mg boosting bolusbefore initializing the PCA protocol [15]. Furthermore, our institu-tional policy is that patients who are experiencing unacceptablepain, are given loading bolus doses of 1–2 mg/min morphine untilpain is relieved by PCA [10,18]. Based on our group’s experience,neither the pre-study booster nor the scheduled rescue drug iscapable of optimally controlling post-colorectal pain for 48 h with-out morphine administration. Thus, by circumventing the abovetwo elements of pain control, we tended at establishing a pharma-cological “stable state”, so it was possible assessing the changeablevariable at stack, i.e., the contribution of a “sub-analgesic” infusionof morphine to analgesic requirements.

The rate of PCA use reflects patient’s pain intensity. The reliabil-ity of subjective scoring of pain and assessment of antinociceptiveefficacy by means of a visual analogue scale (VAS) has been calledinto doubt [19]. The use of VAS has a limitation by representingonly one aspect of pain at the moment of assessment, i.e., inten-sity, given that pain in general confers a variety of sensations. We(unpublished results) and others [20] have noted that VAS wouldchange sporadically at brief intervals for no obvious reason, whilePCA-delivered morphine remains constant. As in our patients, VASscores (or NRS) change little with the initial incremental PCA doses,but they later decrease rapidly to low values (≤3/10 in the BI group).In other words, during the IV-PCA titration, the VAS score does notchange markedly until the morphine dose approached the level

that provides pain relief (the threshold dose). Finally, Auburn et al.reported that nurses used VAS as a guide for pain anti-nociceptionin only 53% of pain assessments [21]. These findings support ourdecision, also backed by our institutional common practice, not to

1 ical R

fpaaDn

mrtptt

b1dnmheaoowTtautatoaaaip

iuri[swatoanab(twvaalggdgfo

90 I. White et al. / Pharmacolog

ollow hourly VAS score as a guide for judging the effect of PCA onain. The PCA effective application (the application of the devicet the unlocked period, at which time it will deliver a bolus of thenalgesic [22]) was utilized to determine the extent of pain control.espite the above, pain scorings were still collected every 12 h, butot as the indicator for analgesic efficacy of the morphine regimen.

The label “sub-analgesic dose” that we used to describe theinimal dose of morphine delivered during infusion, is not a

ecognized pharmacological term. A real-life application of thiserminology is the current dosage of ketamine used for analgesicurposes [10–14]. Similarly, we used the term “sub-analgesic”o describe a protocol where 0.7 mg morphine was deliveredhroughout 1 h for 48 h.

The BI protocol delivered effective analgesia even though theolus dose was 50% less than the “high” dose (i.e., ∼0.7 vs..5 mg/bolus/7 min). Had we combined the infusion with the “high”ose bolus, adverse events may have occurred, and we may haveot answered our principal query regarding the analgesic effect oforphine continuous infusion, because such a combination could

ave provided a complete abolition of pain, thus shadowing theffect of the infusion. Unlike another study [23], where there was

great variability in the morphine requirements of the patients,urs depicted small standard deviations from the mean amountf the self-administered morphine. Our study patients also under-ent the same operation and were treated by the same teams.

he overall optimal results of our study are probably related tohe combination of diclofenac administered at scheduled intervalsnd the PCA protocols. In similar studies, no rescue drugs weresed to coadjuvate morphine, resulting in high morphine consump-ion required to better control pain. This led to higher rates ofdverse events as well, while we obviated such occurrences byhe combination of NSAIDs and morphine. The close monitoringf the patients at the beginning of the study in our PACU, whenll patients were kept for 3-h observation, as well as the use of

strict morphine dosage protocol based on weight adjustment,re both crucial elements in reducing risky events, thus increas-ng the success of this modality of pain management in any patientopulation.

Our findings are inconsistent with previous clinical experimentsn which the background infusion increased postoperative opioidsage [24–26] or the risk of adverse events [23–27], especiallyespiratory depression [28,29]. PCA plus background morphinenfusion has been shown to improve pain relief in earlier studies30,31], but was also associated with an increase in opioid-relatedide effects, such as PONV [31]. Significant respiratory depressionas not observed. Other studies reported that background infusion

dded to the PCA regimen [32–34] increased morphine consump-ion with no improvement in analgesia. In children, the inclusionf a background infusion of 4 �g/kg/h morphine (40% our dose) in

PCA bolus regimen, neither augmented the total morphine usageor increased the frequency of side effects, but was rather associ-ted with less hypoxemia and a better sleep pattern than with noackground infusion [24]. Contrarily, a 5-times larger infusion dose20 �g/kg/h, twice the dose herein used), in a similar population, ledo a higher morphine usage, and side effects (PONV, hypoxemia),ith similar pain scores noted in the background infusion group

ersus the PCA bolus-only one [23]. The latter results support ours,nd imply that the addition of continuous morphine infusion within

given dose range, provides better analgesia without inducing aarger frequency of side effects, or even opioid-induced hyperal-esia (OIH) or tolerance to opioids [15,35], as hypothesized by ourroup prior to setting up the protocol. The variations among the

ifferent reports and ours, might be due to their cohort ethnic ori-in, sex, age and co-morbidities, their relatively small sample sizeor detecting small differences, different background infusion ratesr agents, different types of surgery that induce various degrees of

esearch 66 (2012) 185– 191

pain [36], too-short a duration of PCA use to establish efficacy, ordue to failure to evaluate PCA device applications.

There were two distinctive benefits of the BI protocol comparedto the B0 protocol. First, the total amount of morphine used in theformer group during the study period was ∼25% less than in the lat-ter, even when combining the total bolus dosage with the amountof the infused morphine. Second, the relatively less frequent useof the PCA device by the BI patients (by ∼19%) is indicative of aneuro-pharmacological antinociceptive effect of the low-dose pro-tocol. The much lower rates of application of the device duringthe 24–48 h period compared to the 0–24 h ones reflect the effectachieved by the infused morphine over time in the BI group. Takentogether these and the better pain scores in the BI group, the higherrate of satisfaction and the lack of adverse events, all testify to theantinociceptive pharmacological efficacy and safety of the BI testedinfusion protocol.

The desire to minimize intestinal adverse events, especially afterlarge bowel surgery, is a particularly apt reason to strive for optimalanalgesia by means of as low amounts of opioids as possible. Themost frequent gastrointestinal side effects are dry mouth, constipa-tion, and biliary tract spasm; urinary retention is also related to themorphine-affected genitourinary tract. None of these effects wereexacerbated in the BI group despite the use of continuous infusion.We also consider that the lower amounts of total morphine usecontributed to the BI patients having comfortably left their bedsearlier than the B0 patients.

5. Conclusions

The usual regimen for controlling postoperative pain follow-ing colorectal surgery consists of PCA-bolus-only regimen. In anattempt to lower the total amount of morphine consumption whilepreserving the pain-reducing benefits of its administration, wecombined minimal (half the) doses of PCA morphine boluses (0.7vs. 1.5 mg) with a “sub-analgesic” infusion (0.7 mg/h) dose of mor-phine. This low-dose bolus plus “sub-analgesic” infusion protocolwas more efficacious and satisfactory, and as safe in controllingthe 48-h postoperative pain in the colorectal patients as the PCAbolus-only (1.5 mg) protocol.

Conflict of interest

The authors of this manuscript declare that there are no conflictsof interest.

Acknowledgment

Esther Eshkol is thanked for editorial assistance and Dr. EyalAmar for statistical aid.

References

[1] Poe-Kochert C, Tripi PA, Potzman J, Son-Hing JP, Thompson GH. Continuousintravenous morphine infusion for postoperative analgesia following posteriorspinal fusion for idiopathic scoliosis. Spine [Phila Pa 1976] 2010;35:754–7.

[2] van Dijk M, Bouwmeester NJ, Duivenvoorden HJ, Koot HM, Tibboel D, PasschierJ, et al. Efficacy of continuous versus intermittent morphine administrationafter major surgery in 0–3-year-old infants; a double-blind randomized con-trolled trial. Pain 2002;98:305–13.

[3] Lejus C, Tissot M, Roquilly A, Populaire C, Guillaud C, De Windt A, et al. Riskfactors of postoperative vomiting in children receiving continuous intravenousmorphine. Minerva Anestesiologica 2012;78:160–7.

[4] Aubrun F, Mazoit JX, Riou B. Postoperative intravenous morphine titration.British Journal of Anaesthesia 2012;108:193–201.

[5] Aubrun F, Monsel S, Langeron O, Coriat P, Riou B. Postoperative titration ofintravenous morphine in the elderly patient. Anesthesiology 2002;96:17–23.

[6] Oifa S, Sydoruk T, White I, Ekstein MP, Marouani N, Chazan S, et al. Effectsof intravenous patient-controlled analgesia with buprenorphine and mor-phine alone and in combination during the first 12 postoperative hours:

ical R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

than post-incisional single dose epidural morphine. Minerva Anestesiologica2011;77:408–17.

I. White et al. / Pharmacolog

a randomized, double-blind, four-arm trial in adults undergoing abdominalsurgery. Clinical Therapeutics 2009;31:527–41.

[7] Sucato DJ, Duey-Holtz A, Elerson E, Safavi F. Postoperative analgesia followingsurgical correction for adolescent idiopathic scoliosis: a comparison of continu-ous epidural analgesia and patient-controlled analgesia. Spine 2005;30:211–7.

[8] Dolmatova EV, Imaev AA, Lubnin AY. ‘Scheduled’ dosing of lornoxicam pro-vides analgesia superior to that provided by ‘on request’ dosing followingcraniotomy. European Journal of Anaesthesiology 2009;26:633–7.

[9] Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Gavaghan DJ, et al.Assessing the quality of reports of randomized clinical trials: is blinding nec-essary. Controlled Clinical Trials 1996;17:1–12.

10] Ekstein MP, Weinbroum AA. Immediate postoperative pain in orthopedicpatients is more intense and requires more analgesia than in post-laparotomypatients. Pain Medicine 2010;12:308–13.

11] Weinbroum AA. A single small dose of postoperative ketamine providesrapid and sustained improvement in morphine analgesia in the presence ofmorphine-resistant pain. Anesthesia and Analgesia 2003;96:789–95.

12] Weinbroum AA. Superiority of postoperative epidural over intravenouspatient-controlled analgesia in orthopedic oncologic patients. Surgery2005;138:869–76.

13] Kollender Y, Bickels J, Stocki D, Maruoani N, Chazan S, Nirkin A, et al. Subanaes-thetic ketamine spares postoperative morphine and controls pain better thanstandard morphine does alone in orthopaedic-oncological patients. EuropeanJournal of Cancer 2008;44:954–62.

14] Nesher N, Marouani N, Serovian I, Ekstein MP, Chazan S, Weinbroum AA.Ketamine spares morphine consumption after transthoracic lung and heartsurgery without adverse hemodynamic effects. Pharmacological Research2008;58:38–44.

15] Chen WH, Liu K, Tan PH, Chia YY. Effects of postoperative backgroundPCA morphine infusion on pain management and related side effects inpatients undergoing abdominal hysterectomy. Journal of Clinical Anesthesia2011;23:124–9.

16] Lötsch J, Dubziak R, Freynhagen R, Marschner J, Geisslinger G. Fatal respiratorydepression after multiple intravenous morphine injections. Clinical Pharma-cokinetics 2006;45:1051–60.

17] Umamaheswara Rao GS, Gelb AW. To use or not to use: the dilemma of NSAIDsand craniotomy. European Journal of Anaesthesiology 2009;26:625–6.

18] Weinbroum AA. Non-opioid IV adjuvants in the perioperative period: pharma-cological and clinical aspects of ketamine and gabapentinoids. PharmacologicalResearch 2012;(January) [Epub ahead of print].

19] Leino KA, Kuusniemi KS, Lertola KK, Olkkola KT. Comparison of four pain scalesin patients with hip fracture or other lower limb trauma. Acta AnaesthesiologicaScandinavica 2011;55:495–502.

20] Rovetta G, Monteforte P, Molfetta L. Evaluating pain in osteoarthritis of the

hands: the effect of patient information. International Journal of Clinical Phar-macology Research 2003;23:61–7.

21] Aubrun F, Paqueron X, Langeron O, Coriat P, Riou B. What pain scales do nursesuse in the postanaesthesia care unit. European Journal of Anaesthesiology2003;20:745–944.

[

esearch 66 (2012) 185– 191 191

22] Rakhman E, Shmain D, White I, Ekstein MP, Kollender Y, Chazan S, et al.Repeated and escalating preoperative subanesthetic doses of ketaminefor postoperative pain control in patients undergoing tumor resection:a randomized, placebo-controlled, double-blind trial. Clinical Therapeutics2011;33:863–73.

23] Doyle E, Robinson D, Morton NS. Comparison of patient-controlled analge-sia with and without a background infusion after lower abdominal surgeryin children. British Journal of Anaesthesia 1993;71:670–3.

24] Doyle E, Harper I, Morton NS. Patient-controlled analgesia with low dose back-ground infusions after lower abdominal surgery in children. British Journal ofAnaesthesia 1993;71:818–22.

25] Russell AW, Owen H, Ilsley AH, Kluger MT, Plummer JL. Background infusionwith patient-controlled analgesia: effect on postoperative oxyhaemoglobinsaturation and pain control. Anaesthesia and Intensive Care 1993;21:174–9.

26] Krenn H, Oczenski W, Jellinek H, Krumpl-Ströher M, Schweitzer E, Fitzger-ald RD. Nalbuphine by PCA-pump for analgesia following hysterectomy: bolusapplication versus continuous infusion with bolus application. European Jour-nal of Pain 2001;5:219–26.

27] Macintyre PE. Safety and efficacy of patient-controlled analgesia. British Journalof Anaesthesia 2001;87:36–46.

28] Sidebotham D, Dijkhuizen MR, Schug SA. The safety and utilization ofpatient-controlled analgesia. Journal of Pain and Symptom Management1997;14:202–9.

29] Schug SA, Torrie JJ. Safety assessment of postoperative pain management by anacute pain service. Pain 1993;55:387–91.

30] McKenzie R, Rudy T, Tantisira B. Comparison of PCA alone and PCA with con-tinuous infusion on pain relief and quality of sleep. Anesthesiology 1990;73:A787.

31] Sinatra R, Chung KS, Silverman DG, Brail SJ, Chung J, Harrison DM, et al. Anevaluation of morphine and oxymorphone administered via patient-controlledanalgesia [PCA] or PCA plus basal infusion in postcesarean delivery patients.Anesthesiology 1989;71:502–7.

32] Parker RK, Holtmann B, White PF. Effects of a nighttime opioid infusion withPCA therapy on patient comfort and analgesic requirements after abdominalhysterectomy. Anesthesiology 1992;76:362–7.

33] Owen H, Szekely SM, Plummer JL, Cushnie JN, Mather LE. Variables of patient-controlled analgesia: 2. Concurrent infusion. Anaesthesia 1989;44:11–3.

34] Wu MYC, Purcell GJ. Patient-controlled analgesia—the value of a backgroundinfusion. Anaesthesia and Intensive Care 1990;18:575–6.

35] Bronstein I, White I, Ekstein MP, Brill S, Chazan S, Ogorek D, et al. Higherpostoperative pain and increased morphine consumption follow pre- rather

36] Ho ST, Wang JJ, Huang JC, Lin MT, Liaw WJ. The magnitude of acute tolerance tomorphine analgesia: concentration-dependent or time-dependent. Anesthesiaand Analgesia 2002;95:948–51.