Regional anaesthesia and quality of recovery after surgery

Recovery from surgery is a complex process dependent on surgical, anaesthetic and patient factors, among others. Traditionally, recovery has been measured primarily by physiological parameters and adverse events such as morbidity and mortality 1. More recently, there has been a shift of emphasis towards defining recovery from a patient's perspective, with self-reported measures of recovery incorporating multiple postoperative domains 2. This reflects both a decrease in overall surgical morbidity and mortality due to surgical and anaesthetic advances as well as an increased focus on factors of importance to the patient 3. In this issue of Anaesthesia, Koning et al. report on the impact of an anaesthetic intervention (intrathecal bupivacaine and morphine added to general anaesthesia) on the quality of recovery after robot-assisted radical prostatectomy 4. The primary outcome was the 15-item Quality of Recovery (QoR-15) score and a statistically significant improvement was reported in this study 5. The QoR-15 is a patient-reported outcome measurement of postoperative quality of recovery, published and available on a free-to-access basis 5. This tool was developed by the same group that produced the nine-item ‘QoR’ score in 1999 6, and expanded on it with the longer and more comprehensive QoR-40 in 2000, which demonstrated superior validity and reliability 7. Quality of Recovery-40 has been extensively used and validated in the intervening period 8. Five dimensions are included in QoR-40: pain; physical comfort and physical independence (grouped as physical well-being); and psychological support and emotional state (grouped as mental well-being). Quality of Recovery-15 was in turn derived from QoR-40 as a simplified, shorter version limited to one page which enhances its feasibility and likelihood of completion. Quality of Recovery-15 has equivalent psychometric properties to QoR-40 with an estimated time to completion of 2.5 min vs. 10 min for QoR-40 5. Each of the 15 questions included in this questionnaire is answered on an 11-point numerical rating scale (0–10). The resulting composite score (scaled from 0 to 150) is attained by summing the values from each question to produce a single number. The maximum score of 150 would represent a ‘perfect’ condition with no pain, no nausea or vomiting, no sleep disruption, feeling of well-being, able to perform own hygiene and no anxiety or sadness. Quality of Recovery-15 has been validated in patients undergoing surgery and general anaesthesia across a range of minor, intermediate and major surgical procedures, thus it represents an appropriate measure for quality of recovery in the surgical context of robot-assisted radical prostatectomy. This scoring system has been shown to be reliable, consistent, responsive and clinically acceptable 5, with translation and validation into multiple languages and has undergone systematic review 9. The QoR-15 score (alongside the nine-item QoR score) has recently been recommended by the standardised endpoints in peri-operative medicine initiative as one of six endpoints which should be considered for inclusion in clinical trials assessing patient comfort and pain after surgery 10. Furthermore, the European Society of Anaesthesia (ESA) – European Society of Intensive Care Medicine joint taskforce has recommended QoR-15 as the standardised outcome measure for quality of recovery 11. The use of such validated, standardised measures in clinical trials is deemed necessary to enable accurate comparison of results across studies. A limitation of composite scoring is that it assigns equal value to different recovery parameters, so that, for example, severe pain produces equal weighting to the inability to look after personal hygiene 12. The score is intended for use in clinical trials and in quality assurance, but not at the individual level to determine adequacy of recovery. Several alternative multidimensional quality of recovery assessment scores are available. The Surgical Recovery Index (SRI) 13, Postoperative Quality of Recovery Scale (PQRS) 14 and Surgical Recovery Scale (SRS) 15 have also been validated for inpatient surgical populations. In selecting a suitable scoring system to examine the impact of a regional anaesthetic intervention, pain should form part of the assessment and the time-frame should be the early postoperative period. In this respect, the Surgical Recovery Index (SRI) with a first measurement at postoperative day 7 and the SRS with measurements at baseline, days 3, 7, 30 and 60 would not be ideally suited, as they may miss the early recovery phase, the window during which regional anaesthesia consistently has the greatest impact on recovery. The PQRS tool is based on six domains of recovery, including nociception, with a dichotomous scoring system in which the postoperative return to a patient's previous baseline is assessed to determine adequacy of recovery. This system provides flexibility in the assessment of each domain separately and includes a cognitive assessment with a range of timeframes from the immediate postoperative period through days 1, 3 and at 3 months if required 14. The PQRS score, therefore, could also be considered to assess the effect of a regional anaesthesia intervention. In this context, we commend Koning et al. for the use of QoR-15 as a primary outcome measure for their study. As is recommended, the questionnaire was conducted before surgery to provide baseline data and then repeated at 24 h after surgery 11. The premise that improvement in analgesia due to the study intervention would lead to a significantly smaller decline in QoR-15 score would appear to be a reasonable and indeed intuitive expectation. While only two of the QoR-15 questions explicitly refer to pain, other questions such as to what extent the patient had been ‘feeling comfortable and in control’ and to what extent they ‘had a good sleep’ might be expected to involve pain as a factor. A question of paramount importance when interpreting the results of a trial is what is the difference in score that would be considered minimally clinically important; in other words, significant enough to warrant the intervention? Smaller differences below that threshold would be considered clinically negligible and would not support the intervention, irrespective of the result of statistical tests. In the study by Koning et al., the difference in QoR-15 score between the study and control groups was modest (5 points on the scale of 0–150). As the authors have acknowledged, this is less than the 8 points defined as the MCID 16, and on this basis, we would question if it is accurate to describe the outcome of this study as a positive result. Although the difference of 5 points was statistically significant, the magnitude of the difference is likely clinically negligible 17. When the five subdomains of QoR-15 were individually analysed, the domain of ‘pain’ (two questions on the extent of moderate pain and severe pain over the preceding 24 h) was the only one of the five to yield statistical significance. Within this domain, the effect size was relatively small with 6/20 in the control group vs. 2/20 in the intervention group. The postoperative morphine consumption via patient-controlled intravenous analgesia was 1.5 mg in the study group vs. 5 mg in the control group, again a small difference of questionable clinical significance. The striking finding is the relatively low pain scores and opioid consumption recorded in the control group receiving standard care in this centre. Consequently, the scope for improvement in quality of recovery by an intervention addressing acute pain would naturally be expected to be limited. Although the study was powered based on an estimated decline in QoR-15 at postoperative day 1 (POD 1) of 35% in the control group and 25% in the study group, the actual decline in the control group was only about 12% (or 18 points). In fact, the score for the control group on POD1 was 118, which coincidentally is the threshold previously defined as the ‘patient acceptable symptom state’ 16. The ‘acceptable’ recovery score in the control group by virtue of the efficacy of multimodal analgesia has made further significant clinical improvement from the addition of intrathecal morphine challenging to achieve. In order to use QoR-15 to its maximum potential we need to establish when its use as a research tool is most appropriate. The impact of clinical trials is dependent on the use of well-defined patient-centred outcome measures that are potentially modifiable by the trial intervention 11. It stands to reason that more invasive surgical procedures affecting major organ systems, with high levels of postoperative pain and functional limitation, will have the greatest postoperative declines in QoR-15, and in turn will be more susceptible to improvement from an effective intervention. Indeed, Myles et al. have reported 24-h mean postoperative scores of 127, 114 and 106 for minor, intermediate and major surgery, respectively 16. At the same time, major regional anaesthesia interventions with substantial positive impact on several aspects of physiology (e.g. thoracic epidural analgesia for major laparotomy) 18 are more likely to improve postoperative outcomes than more minor regional anaesthesia interventions aimed at improving analgesia as part of a multimodal regimen (e.g. an adductor canal block for total knee arthroplasty). In terms of future clinical trial designs, it may be prudent to undertake a pilot study to determine an accurate baseline Q0R 15 in the control (standard care) group. If a small postoperative decline is noted with the current standard of care, it may be unlikely that any intervention will produce an improvement in individual domains of recovery large enough to yield a positive result (of at least 8 points). In addition, an isolated intervention that in itself is expected to provide a small, incremental benefit (e.g. postoperative paracetamol), although unlikely to improve QoR 15 in and of itself, may be a useful component of a larger multimodal strategy. Mature multimodal analgesic regimens that include regional anaesthesia and systemic analgesics are gold-standard therapy for many surgical procedures, and when taken as a group, have been proven to effectively treat pain and reduce side-effects 19. It remains necessary to assess if the addition of an analgesic intervention to a ‘best-practice’ multimodal technique further improves pain control or allows replacement of another analgesic intervention to improve cost effectiveness and/or safety 20. However, it is unlikely that if studied in isolation each individual component of the regimen would result in a significant improvement to a complex outcome such as QoR-15. This begs the question of how worthwhile new interventions are in clinical scenarios where an ‘acceptable’ baseline quality of recovery already exists. As with any treatment, the potential benefits must be weighed against the clinical risks as well as cost and resource requirements, within a cost effectiveness framework. In the context of marginal gains in patient outcomes, this presents a challenge as we strive to innovate and improve standards of care. In certain outcomes, such as the goal of reducing opioid consumption, some benefits of novel techniques and treatments may not necessarily be reflected in the QoR-15 score if recovery from the patient's perspective is not altered significantly. In retrospect, had pain score been examined as a primary instead of secondary outcome in this study it is a reasonable assumption that statistical significance would also have been demonstrated. It would be overly speculative to comment on whether any difference in this outcome might have been clinically significant in this instance. Regardless, a clearly defined and accepted clinically important difference (i.e. two points in an eleven point numerical rating scale) should be demonstrated to support positive findings in a pain score outcome 21. Traditional outcome measures such as pain or opioid consumption represent a more conservative option for investigators, whereas a positive finding on a multidimensional quality of recovery scale would carry greater clinical weight. For this reason, we support a greater use of QoR15 for major regional anaesthetic techniques expected to result in significant pain relief, improvement in sleep quality and overall sense of well-being after a relatively invasive and painful surgical procedure (e.g. continuous brachial plexus analgesia for total shoulder replacement). For a minor regional anaesthesia intervention expected to provide an incremental improvement for a minimally invasive surgical procedure (e.g. a chest wall block as part of a multimodal regimen for minor breast surgery), more traditional outcome measures such as pain relief and opioid consumption are still reasonable outcomes to study. In summary, the study by Koning et al. highlights the value of the QoR-15 in providing a standardised, validated measure of postoperative recovery. In this respect, the described intervention demonstrated a marginal improvement in pain management specifically and quality of recovery more broadly, but not to a clinically meaningful extent across the study population. In general terms, the impact on quality of recovery will be most evident with important anaesthetic interventions targeted at major surgical procedures that usually result in severe pain, nausea and vomiting and significant functional limitation. Quality of Recovery-15 provides a levelling mechanism, creating an objective measurement to compare improvements in one or more individual domains within the broader context of patient-centred outcomes. In this standardised manner, the relative value of new interventions may be more readily interpreted by the wider anaesthetic community. AP has a current research grant from Fisher and Paykel and she receives protected academic time from the Department of Anesthesia, University Health Network, Toronto. No other competing interests declared.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.