Revision of expert panel’s guidelines on postoperative pain management

The French Society of Anaesthesia and Intensive Care Medicine (SFAR) published experts’ guidelines on the care of postoperative pain. This was an update of the 2008 guidelines. Fourteen experts analysed the literature (PubMed, Cochrane) on questions that had not been treated in the previous guidelines, or to modify the guidelines following new data in the published literature. The used method is invariably the GRADE C method, which guarantees a rigorous work. Seventeen recommendations were formalised on the assessment of perioperative pain, and most particularly in non-communicating patients, on opioid and non-opioid analgesics and on anti-hyperalgesic drugs, such as ketamine and gabapentinoids, as well as on local and regional anaesthesia. The concept of vulnerability and therefore the identification of the most fragile patients in terms of analgesics requirements were specified. Because of the absence of sufficient data or new information, no recommendation was made about analgesia monitoring, the procedures for the surveillance of patients in conventional care structures, or perinervous or epidural catheterism. C 2019 The Authors. Published by Elsevier Masson SAS on behalf of Société française d’anesthésie et de réanimation (Sfar). This is an open access article under the CC BY license (http://creativecommons.org/

The French Society of Anaesthesia and Intensive Care Medicine (SFAR) published experts' guidelines on the care of postoperative pain. This was an update of the 2008 guidelines. Fourteen experts analysed the literature (PubMed TM , Cochrane TM ) on questions that had not been treated in the previous guidelines, or to modify the guidelines following new data in the published literature. The used method is invariably the GRADE C method, which guarantees a rigorous work. Seventeen recommendations were formalised on the assessment of perioperative pain, and most particularly in non-communicating patients, on opioid and non-opioid analgesics and on anti-hyperalgesic drugs, such as ketamine and gabapentinoids, as well as on local and regional anaesthesia. The concept of vulnerability and therefore the identification of the most fragile patients in terms of analgesics requirements were specified. Because of the absence of sufficient data or new information, no recommendation was made about analgesia monitoring, the procedures for the surveillance of patients in conventional care structures, or perinervous or epidural catheterism.

Introduction
Since the consensus conference on postoperative pain (POP) in 1997 and the experts' guidelines of 2008, it was necessary to complete or modify the existing guidelines. A group of 14 experts has worked on questions that have not been treated in the previous guidelines, or on published guidelines to be modified following new data in the literature. The method used remains the GRADE method, which guarantees a rigorous work.

Methodology
The method used to elaborate these guidelines was the GRADE 1 method. Following a quantitative literature analysis, this method was used to separately determine the quality of available evidence, i.e. estimation of the confidence needed to analyse the effect of the quantitative intervention, and the level of recommendation. The quality of evidence was rated as follows: high-quality evidence: further research is very unlikely to change the confidence in the estimate of the effect; moderate-quality evidence: further research is likely to have an impact on the confidence in the estimate of the effect and may change the estimate of the effect itself; low-quality evidence: further research is very likely to have an impact on the confidence in the estimate of the effect and is likely to change the estimate of the effect itself; very low-quality evidence: any estimate of the effect is very unlikely.
The analysis of the evidence's quality is performed for each study. A global level of evidence is then defined for a given question and criterion. The final wording of the recommendations will always be binary -either positive or negative, either weak or strong: strong recommendation: we recommend or we do not recommend (GRADE 1+ or 1À); weak recommendation: we suggest or we do not suggest (GRADE 2+ or 2À).
The strength of the recommendation is determined by key factors, and approved by the experts after a vote, using the Delphi and GRADE grids: estimate of the effect; global level of evidence: the higher this level, the stronger the recommendation; the balance between desirable and adverse effects: the more favourable the balance, the higher the recommendation; values and preferences: in the event of uncertainties or great variability, the recommendation will most likely be weak. These values and preferences must be obtained directly from the people involved (patient, doctor, decision-maker); costs: the higher the costs or the use of resources, the weaker the recommendation; to develop a recommendation, at least 50% of participants must have an opinion and less than 20% the opposite opinion; to develop a strong recommendation, at least 70% of participants must agree (grade between 7 and 10).
For some questions, the existence of several studies and/ or meta-analyses of acceptable methodological quality, the GRADE method applied entirely and allowed making recommendations.
If experts did not have a meta-analysis allowing them to respond to the question, a quantitative analysis following the GRADE method was possible, and a systematic review was performed. An expert opinion was then proposed and approved if at least 70% of experts agreed with the proposal.
Finally, in some fields, the lack of recent studies did not allow us to formulate recommendations.
The questions tackled in these Guidelines updates are the following: After the synthesis of our experts' work and the implementation of the GRADE method, 17 recommendations were formalised by the organisational committee. Among these, 11 are strong, 3 are weak and for 3 recommendations, the GRADE method could not be applied, and they are therefore formulated as experts' opinions.
The entirety of these recommendations was submitted to a proofreading group for a Delphi type rating. After two rounds of scoring and several amendments, a strong agreement was reached for 14 recommendations.
Pain assessment during the perioperative period 1. When and why to assess it? Benefits of the assessment on postoperative consequences and chronicisation, 2. Should the DN4 form be used perioperatively?
R1.1 -During the preoperative period, we recommend identifying the most vulnerable patients to pain (with risks to develop severe postoperative pain and/or a chronic post-surgical pain (CPSP)) by focusing on preoperative pain, including pain even far from the operating site, the long-term consumption of opioids, and surgical and psychological factors such as anxiety or depression. Experts' opinion, Strong agreement Argument: identifying patients implies a cautious supervision with a multimodal therapeutic strategy, including, whenever possible, regional analgesia and the administration of anti-hyper analgesic drugs.
Surgical factors are: surgical procedures such as thoracotomy [1], breasts surgery, sternotomy and iliac crest sampling [2,3], most likely to cause CPSP; repeated surgical procedures causing a risk of CPSP higher than the initial surgery [4], because of the high number of nerve damages on a revised scar tissue, more inflammatory tissues; surgical procedures with preoperative pain on the site of the intervention a duration of surgery greater than 3 hours.

R1.2 -
We recommend using the APAIS scale (Amsterdam Preoperative Anxiety and Information Scale) as a measure anxiety and/or the need for information during the preoperative period. Experts' opinion, Strong agreement Argument: anxiety, stress and depression are the psychological factors most likely to cause severe postoperative pain [5][6][7]. They play an important role for CPSP development. Catastrophism is a predictive factor of a more intense postoperative pain, of greater morphine consumption in various surgical models, but also of a more frequent CPSP in orthopaedic surgery. The assessment during the pre-anaesthesia evaluation, using a simple scale (APAIS) would allow predicting the transition from acute to chronic POP.
R2. -We recommend identifying the postoperative risk factors for CPSP by searching a high intensity of POP (using a numerical scale), an uncommon prolongation of POP, an early neuropathic pain (using a DN4 scale), or signs of anxiety or depression.
Experts' opinion, Strong agreement Argument: An early neuropathic pain should be properly treated [3]. The result of a positive screening of neuropathic pain should be reported to the patient, the surgeon and the referring physician.
3. Which pain scale(s) to use in children under 7 years old?
R3. -We recommend using a self-assessment scale from 5 years old (face scale, Appendix 1). Otherwise, we recommend using the FLACC scale for postoperative pain assessment in children under 7 years of age (Appendix 2).

G1+, Strong agreement
Argument: for new-borns: the EDIN (neonatal pain and discomfort scale), DAN (acute neonatal pain) and NFCS (neonatal facial coding system) scales can be used but have not been validated for postoperative pain. The FLACC (face legs activity cry consolability) scale can be used from 2 months old [8][9][10]. EVEN-DOL is a pain scale approved for an outside hospital use and Emergency unit only [11]. As for the EVA pain scale, it must be presented vertically to the child. 4. Which scale(s) to use in non-communicating patients?
R4. -For communicant patients, we recommend using a modified FLACC scale in children, and ALGOPLUS in elderly people.
G1+, Weak agreement Argument: The modified FLACC scale can be used in noncommunicant patient from birth until 18 years old and contains 5 simple behavioural items: face, legs, activity, screams and consolability [12,13]. For the ALGOPLUS scale [14], a grade ! 2/5 allows the diagnosis of pain with an 87% sensibility and a 80% specificity.
5. What are the procedures for monitoring the patients under opioid-like drugs in postoperative period?
The procedures for monitoring the patients under opioid-like drugs by subcutaneous route, by patient-controlled analgesia (PCA) or by epidural route were specified in the consensus conferences on postoperative pain management in adults and children of 1997 and 1999. These do not need to be changed.

No recommendation
Argument: since these last two recommendations, high-risk patients for respiratory depression-related to opioids were identified [15][16][17][18]. The literature highlights risks: patient over 70 years old, first-time on opiates, morbid obesity (BMI > 35), respiratory disease, obstructive sleep apnoea (OSA), patient with liver or kidney failure, or describing an intense pain suddenly stopping, the association of opiates and drugs that can cause a depression of the central nervous system such as benzodiazepines, barbiturates, antidepressants, antiemetics or antihistaminic drugs. Likewise, the association of opiates with alcohol or illegal drugs, a history of neurological and/or neuromuscular disorders or the perimedullary route [19,20] require a reinforced monitoring, including a pulse oximetry.
A more frequent clinical and/or non-invasive monitoring (plethysmography, capnography in the post-anaesthesia care unit) are probably suggested for high-risk patients, and more specifically in case of OSA and for strong doses of opiates/sedatives consumed in perioperative period.
Apart from these various reminders, no new data enables to write more precise recommendations.
Analgesia monitoring: what are the methods enabling a monitoring of analgesia in the immediate postoperative period in children and adults?

No recommendation
Argument: the analysis of literature established that pupillometry, the ANI (Analgesia Nociception Index) and the SPI (Surgical Pleth Index) allow the correct assessment of the analgesianociception balance under general anaesthesia [21][22][23]. However, it is at the moment impossible to assert that such perioperative monitoring enables a decrease in postoperative pain or in the consumption of postoperative analgesics.
In awake patients, some publications suggest a degree of correlation between the ANI or the pupillometry and the pain scores obtained from the self or hetero-assessment scales [24,25]. However, the use of these monitors in the immediate postoperative period did not prove to be better or more beneficial than auto or hetero-assessment scales.
It is however important to note that no study was performed in postoperative non-communicant patients, population for which pain assessment scales are most difficult.
Apart from these various reminders, no new data enables to write more precise recommendations 9. Drug therapies via systemic and oral routes 6. What is the place for selective and non-selective NSAIDs in perioperative?
R6.1. -We recommend the association of a non-selective NSAID (NS-NSAID) or a selective inhibitor of type 2 cyclooxygenase (ISCOX2) with morphine if there is no contraindication to the use of NSAIDs.

G1+, Strong agreement
Argument: NS-NSAIDs or ISCOX2, associated with morphine, allow an improvement of pain scores, a significant morphine sparing coupled with a decrease of sedation, PONVs and of the length of postoperative ileus. This association allows the most important decrease in morphine consumption when compare with other associations involving other non-opioid analgesics (nefopam or paracetamol). The level of evidence is high and based on an abundant literature of high methodological quality. Fifteen studies, amongst which 2 meta-analyses, have studied the benefit of NS-NSAIDs associated with morphine, and 25 studies, amongst which 1 meta-analysis, have studied the benefit of ISCOX2 associated with morphine. NS-NSAIDs and ISCOX2 were compared in 6 studies: 4 RCTs and 2 meta-analyses. Results showed similar benefits in terms of analgesia for both drugs either used alone or associated with morphine.
NSAIDs are probably recommended after colorectal surgery but a doubt remains on the risk of anastomotic leakage (Guidelines on enhanced recovery after colorectal surgery) About the renal risk, we do not recommend prescribing a NSAID (NS or ISCOX2) in the event of renal hypoperfusion. An estimated clearance of plasma creatinine below 50 mL/min is a contraindication to NSAID [26,27].
G1-, Strong agreement R6.3 -We do not recommend administering NS-NSAIDs in patients atherothrombosis (PAD (peripheral artery disease), stroke, myocardial infarction) for more than 7 days G2-, Strong agreement Argument: The increased risk of atherothrombotic events associated with a chronic treatment (mainly after 7 days of treatment), with NS-NSAIDs or ISCOX2 is highly documented [28,29,30,31] in the medical literature. In the perioperative period, the atherothrombotic risk associated with ISCOX2 is well documented [32,33]. For NS-NSAIDs, the only two available studies are retrospective but have included a large number of patients (10 873 and 1 309); they did not report an additional cardiovascular risk [34,35].

R6.4 -We do not recommend associating NS-NSAIDs with curative doses of anticoagulant. G1-, Weak agreement
Argument: the NSAIDs used in France (ketoprofen, ibuprofen) in the perioperative period do not increase the risk of postoperative haemorrhage, including after tonsillectomy [36,37]. No study assessing postoperative bleeding as the main outcome has been published so far [38][39][40]. The studies that have highlighted a haemorrhagc risk are either retrospective (n = 1) [41] or a metaanalysis [42] that has included the retrospective studies with an important heterogeneity. All of them studied ketorolac, which is not available in France. According to Bellis et al.'s meta-analysis published in 2014 [43] and including 15 studies and 1693 patient, NSAIDs administered with dexamethasone are not associated with an increased risk of haemorrhage. However, the association of NSAIDs and a curative dose of anticoagulants (enoxaparine, rivaroxaban or VKA) was shown to multiply by 2.5 the risk of severe [44].
7. What place for oxycodone in perioperative: main indications? What limits?
R7. -We recommend to prescribe a strong opiate (morphine or oxycodone), ideally through oral route, for severe postoperative pain or if weaker analgesics are not powerful enough to relief the patients. This recommendation applies to all ages.

G1+, Strong agreement
Argument: morphine remains the reference strong opiate in postoperative period. The oral route must be favoured as much as possible [45,46]. The clinical efficacy of oxycodone equals that of morphine, with a ratio of 1/1 for the IV route and ½ for the oral route (5 mg of oxycodone = 10 mg of morphine sulphate). 8. What are the indications on lidocaine in perioperative? What dose?
R8. -We suggest that adults who undergo major surgery (abdominal, pelvic or spinal surgeries) and who do not benefit regional analgesia receive intravenous lidocaine infusion (bolus: 1 to 2 mg/kg followed by 1 to 2 mg/kg/h) in order to decrease the level of postoperative pain and to improve recovery.

G2+, Strong agreement
Argument: lidocaine is a local anaesthetic usually administered for nerve or epidural block. Intravenous lidocaine presents analgesic, anti-hyperalgesic and anti-inflammatory properties. Comparisons between intravenous lidocaine and loco regional analgesia techniques will allow clarifying the role for each of these techniques. the panel proposed a bolus dose of IV lidocaine 1-2 mg.kg-1 followed by a continuous infusion of 1 to 2 mg/kg-1/h-1 [47][48][49][50]. 9. What place for corticoids in premedication, during and after surgery?
R9. -We suggest that adults receive dexamethasone IV at 8 mg to reduce postoperative pain. G2+, Strong agreement Argument: corticosteroids are widely used in anaesthetised patients. Dexamethasone, given at induction of anaesthesia, is the most studied glucosteroid in anaesthesia and decreases the risk of postoperative nausea and vomiting Its effect on postoperative pain was assessed in numerous studies [51,52]. The recommended dose of dexamethasone is 8 mg in adults and 0.15 mg/kg in children.
10. Should small doses of ketamine be administered to all patients during surgery? What dose? When to administer it (before the incision or before the induction)? Should ketamine be used in postoperative and if yes, in which patients?
R10. -Intraoperatively, small doses of ketamine in patients on general anaesthesia are recommended in the two following situations: 1/ surgery with high risk of acute pain or chronic postoperative pain; 2/ patients with vulnerability to pain, and most particularly patients taking long-term opioids or addicted to opioids.

G1+, Strong agreement
Argument: ketamine is the anti-hyperalgesic drug recommended in first intention at a maximum dose of 0.5 mg/kg/h after anaesthesia induction (to prevent psychodysleptic side effects) more or less in continuous at a dose between 0.125 and 0.25 mg/ kg/h. Perfusion will be stopped 30 min before the end of the surgery.
The use of ketamine at small doses during the surgical period allows a decrease in acute pain intensity for 24 hours, a mean drop of 15 mg in 24 hours of morphine consumption and a decrease in the risk of nausea and vomiting (moderate evidence) [53,54]. The continuation of ketamine treatment in the postoperative period increases the risk of hallucinations and does not importantly increase the analgesic effect. The effect on chronic postoperative pain is an estimated decrease of 30% in the incidence of chronic pain three months after surgery (low level of evidence) [55,56]. We cannot specify if the prolongation of treatment for 24 hours can allow a decrease in the risk of postoperative chronic pain.
Use of magnesium is not currently recommended because of a too low level of evidence 11. What place for gabapentinoids in pre and postoperative.
R11. -The systematic use of gabapentinoids preoperatively is not recommended for postoperative pain management.

G1-, Weak agreement
Argument: The use of gabapentine or pregabaline in premedication allows decreasing the pain intensity during the first postoperative day. The consumption of morphine and the risk of nausea and vomiting. However, both drugs are responsible for an increased risk of sedation and dizziness, and visual impairment (pregabaline) [55][56][57][58][59]. There is no noticeable effect on the prevention of chronic postoperative pain (high level of evidence) [55,59,60]. If the benefit/risk balance is taken into account, gabapentinoids should not be used systematically, or in outpatient surgery. There is no evidence on the interest of coupling gabapentinoids to ketamine. The key anti-hyperalgesic is ketamine. Patients who seem to benefit more from gabapentinoids in the immediate postoperative period are patients who undergo heavy pro-nociceptive surgeries such as arthroplasties, spinal surgery and amputations (high level of evidence). Gabapentinoids can also bring an interesting preoperative sedation if this effect is desired.
Postoperative local and loco-regional anaesthesia 12. What are the indications and limits for postoperative perinervous catheterisation?
13. What are the indications and limits for an epidural and paravertebral catheterisation?

No recommendation
Argument: recent literature confirms the interest of a perinervous catheter when there is a risk of postoperative pain, moderate to severe, and most particularly in shoulder (interscalenic) or knee (femoral) prosthetic surgery [61,62]. In addition to prolonged analgesia efficiency, the benefits concern the sparing effect of opioids and the decrease in morphine adverse effects (postoperative nausea and vomiting), the improvement of sleep and the satisfaction of patients.
However, the literature cannot demonstrate the input of perinervous catheters on pain chronicisation.
The risk of catheter mobilisation (5 to 25%) potentially decreases the analgesic benefit. The femoral catheter, by motor block induced and prolonged, can foster falls, alter rehabilitation and early recovery after a knee prosthetic surgery. The interscalenic catheter induces a diaphragmatic paresis to take into account in the event of respiratory pathology.
Apart from these various reminders, no new data allows writing more precise recommendations.
14. What are the indication and limits for an infiltration catheterisation?
R14.1 -It is recommended not to exceed the maximal toxic doses of local anaesthetics, most specifically for peri-prosthetic orthopaedic infiltrations and during associations of scar infiltrations and perinervous analgesic catheters. G1+, Accord fort Argument: For information, the maximum doses for the first injection of local anaesthetics in a young adult of class ASA 1 are reminded in the R14.2 -In case of laparotomy (laparotomy, caesarean section and lumbotomy), and in the absence of epidural analgesia, we probably recommend the implementation of a continuous infiltration catheter. G2+, Strong agreement Argument: Numerous infiltration protocols of the surgical site are recommended as an analgesic alternative to peripheral nervous catheters but their efficiency is less important in terms of analgesia after the 24th hour [63][64][65].
R14.3 -We do not recommend performing an analgesic infiltration using an intra-articular catheter because of the toxic risk of local anaesthetics on cartilage.