A core outcome set of measurement instruments for assessing effectiveness and efficacy of perioperative pain management: results of the international IMI-PainCare PROMPT Delphi consensus process

The Innovative Medicines Initiative 2 Joint Undertaking under grant agreement 777500 [to the IMI-PainCare project]. This Joint Undertaking receives support from the European Union's Horizon 2020 research and innovation programme and EFPIA.

The rationale and framework of the COS initiative, part of the EU-funded IMI-PainCare PROMPT project, were designed following COMET recommendations for COS development and the COSMIN methodology for evaluating PROMs for a COS 18 , 19 , 20 , 22 and reported according to Core Outcome Set -STAndards for Reporting (COS-STAR) guidelines. 23 The protocol for developing this COS was registered in the COMET database ( https://www.comet-initiative.org/Studies/Details/1731 ) and published in detail before commencing all associated studies. 14 Building on the initial consensus process ( Fig. 1, I . COS of domains), which established a COS of domains for evaluation of perioperative pain management across clinical trials, observational studies, and clinical practice, 21 this study focused on developing a COS of PROMS for the four domains. The stepwise process ( Fig. 1, II . COS of PROMs) included the following: (1) systematic identification of potential PROMs by the IMI-PainCare PROMPT project steering committee, (2) extensive evaluation of their psychometric properties following COSMIN and COMET recommendations, and (3) a consensus process using a Delphi approach with an international, multiprofessional stakeholder panel, including persons with lived experience (PWLE). Fig 1 Overall study design of the IMI-PainCare PROMPT development process for a core outcome set (COS) of patient-reported outcome measures (PROMs) for the assessment of acute postoperative pain in adults. The COS process began with developing a COS of domains (I, in orange), 21 followed by the development of a COS of PROMs (II, in green; this study). (I) Following SLRs on domains assessed in clinical postoperative pain, the IMI-PainCare PROMPT consensus panel consented on a COS of domains comprising four overarching outcome domains 21 : pain intensity with the subdomains ‘in general’, ‘at rest’, and ‘during activity’; physical function; adverse events; and self-efficacy. (II) Subsequently, SLRs were conducted to identify PROMs potentially suitable to assess pain intensity (in general, at rest, and during activity), physical function (physical function after TKA 11 ), adverse events, and self-efficacy after surgery. The identified instruments were evaluated following COSMIN recommendations (e.g. regarding content validity assessment of scales). Sensitivity-to-change analysis of preselected PROMs was carried out in a prospective multicentre trial. 3 Results of these analyses were presented to the IMI-PainCare PROMPT consensus panel in an online meeting, followed by an online Delphi process and a concluding consensus meeting. Finally, a COS of PROMs was recommended for assessing effectiveness and efficacy of perioperative pain management. Similar to the process on domains, representative types of surgery were selected: BS, ST, TKA, and ES. The reasoning behind choosing these four types of surgery is detailed in a previous publication. 3 , 21 AE, adverse event; BS, breast surgery; COS, core outcome set; ES, surgery related to endometriosis; NIT- 1, non-interventional trial 1; PF, physical function; PI, pain intensity; PROM, patient-reported outcome measure; PROMPT, PROMs for postoperative pain treatment; SE, self-efficacy; SLR, systematic literature search; ST, sternotomy; TKA, total knee arthroplasty. Fig 1 Overall study design of the IMI-PainCare PROMPT development process for a core outcome set (COS) of patient-reported outcome measures (PROMs) for the assessment of acute postoperative pain in adults. The COS process began with developing a COS of domains (I, in orange), 21 followed by the development of a COS of PROMs (II, in green; this study). (I) Following SLRs on domains assessed in clinical postoperative pain, the IMI-PainCare PROMPT consensus panel consented on a COS of domains comprising four overarching outcome domains 21 : pain intensity with the subdomains ‘in general’, ‘at rest’, and ‘during activity’; physical function; adverse events; and self-efficacy. (II) Subsequently, SLRs were conducted to identify PROMs potentially suitable to assess pain intensity (in general, at rest, and during activity), physical function (physical function after TKA 11 ), adverse events, and self-efficacy after surgery. The identified instruments were evaluated following COSMIN recommendations (e.g. regarding content validity assessment of scales). Sensitivity-to-change analysis of preselected PROMs was carried out in a prospective multicentre trial. 3 Results of these analyses were presented to the IMI-PainCare PROMPT consensus panel in an online meeting, followed by an online Delphi process and a concluding consensus meeting. Finally, a COS of PROMs was recommended for assessing effectiveness and efficacy of perioperative pain management. Similar to the process on domains, representative types of surgery were selected: BS, ST, TKA, and ES. The reasoning behind choosing these four types of surgery is detailed in a previous publication. 3 , 21 AE, adverse event; BS, breast surgery; COS, core outcome set; ES, surgery related to endometriosis; NIT- 1, non-interventional trial 1; PF, physical function; PI, pain intensity; PROM, patient-reported outcome measure; PROMPT, PROMs for postoperative pain treatment; SE, self-efficacy; SLR, systematic literature search; ST, sternotomy; TKA, total knee arthroplasty. Systematic literature reviews (SLRs) were conducted to identify potential PROMs for the predefined COS of domains, 21 followed by appraisal of psychometric properties according to the COSMIN methodology. 18 , 19 , 20 , 22 The methods for SLRs are detailed in Supplementary material S1 , including PRISMA flow charts ( Supplementary Figs S1–S7 ) and PICOS criteria ( Supplementary Tables S1–S7 ). The methods for evaluating the psychometric properties of all PROMs suggested for inclusion in the Delphi process are detailed in Supplementary material S1 , Chapter S2.1. The methods of the Delphi process adhered to the ethical principles of the Declaration of Helsinki and Good Epidemiological Practice. The planned consensus process was submitted to the Ethics Committees of Dresden and Muenster, who determined that IRB approval and, consequently, obtaining informed consent, were unnecessary (EK 261052019; 2019-151-b-S, amended 26.08.2022). The PROMPT consensus panel consisted of 40 members, including clinicians, researchers, and stakeholders from pharmaceutical companies, health technology assessment agencies, regulatory bodies, and PWLE, treated equally alongside the experts. Four members were invited per stakeholder group, including those with lived experience. Details on the panel composition and recruitment process are provided in Supplementary material S1 , Chapter S3.1, and in our previous publication. 2 , 21 Thirty-nine panel members agreed to participate in the IMI-PainCare PROMPT Delphi consensus process on PROMs, with 36 voting in any of the Delphi rounds. A total of 33 participants voted in round 1 (including four PWLE), 31 voted in round 2 (including three PWLE), and 33 (including three PWLE) attended the final consensus meeting. Participant characteristics, including age, sex, experience with consensus processes, profession, and other variables, are presented in Table 1 . Table 1 Participant characteristics. Self-reported participant characteristics and classifications for voters in the Delphi process: HP, including PR, HCP, and HTA, and PWLE∗. Total counts can be less than actual n because voters were not required to answer all participant characteristic questions. HCP, clinicians/healthcare professionals; HP, healthcare professionals; HTA, healthcare technology assessment experts; PR, pain researchers; PWLE, persons with lived experience of pain. ∗We retained the abbreviation ‘PWLE’ for the tables to facilitate readability and because it is commonly used in publication including patient representatives; however, upon request from participating individuals who have experienced pain, we refer to them as ‘those with lived experience’ throughout the text. † Previous round: For the first round, this question refers to the previous Delphi process on domains. 21 ‡ Although experts were invited to represent specific roles, many of them held multiple positions, such as being both clinicians and researchers or anaesthesiologists and pain specialists. Table 1 Round 1 Round 2 Overall ( n =33) HP ( n =29) PWLE ( n =4) Overall ( n =31) HP ( n =28) PWLE ( n =3) Sex  Female 15 11 4 15 12 3  Male 17 17 0 16 16 0  Unknown 0 0 0 0 0 0  Prefer not to answer 1 1 0 1 1 0  Other 0 0 0 0 0 0 Voted in previous round †  Yes 22 19 3 28 25 3  No 9 8 1 4 4 0  Do not remember 2 2 0 0 0 0 Invited as ‡  PR 17 17 0 19 18 0  HCP 8 8 0 7 7 0  HTA 2 2 0 3 3 0 Professions (self-reported) ‡  Anaesthesiologists 8 7  Pain specialists 7 4  Surgeons 2 3  Physiotherapist 3 4  Nurses 4 2  Psychologists 3 4  Gynaecologists 0 1  Health economists 1 1  Neurologists 0 1  Clinical development scientist 0 1  Clinical researcher 1 1  Preclinical scientist 1 0 PWLE 4 0 4 3 0 3 Other 2 2 0 0 0 0 Years of clinical experience  10 16 16 0 19 19 0 Years of research experience  10 25 24 1 26 24 2 Years of COS experience  None 4 4 0 2 2 0  10 16 14 2 16 14 2 Region of residence  Belgium 3 3 0 2 2 0  France 1 1 0 1 1 0  Germany 11 5 2 8 7 1  Greece 1 1 0 1 0 0  Iceland 1 1 0 0 0 0  The Netherlands 3 3 0 2 2 0  New Zealand 0 0 0 1 1 0  UK 4 2 2 5 4 1 Participant characteristics. Self-reported participant characteristics and classifications for voters in the Delphi process: HP, including PR, HCP, and HTA, and PWLE∗. Total counts can be less than actual n because voters were not required to answer all participant characteristic questions. HCP, clinicians/healthcare professionals; HP, healthcare professionals; HTA, healthcare technology assessment experts; PR, pain researchers; PWLE, persons with lived experience of pain. ∗We retained the abbreviation ‘PWLE’ for the tables to facilitate readability and because it is commonly used in publication including patient representatives; however, upon request from participating individuals who have experienced pain, we refer to them as ‘those with lived experience’ throughout the text. † Previous round: For the first round, this question refers to the previous Delphi process on domains. 21 ‡ Although experts were invited to represent specific roles, many of them held multiple positions, such as being both clinicians and researchers or anaesthesiologists and pain specialists. To facilitate informed decision-making from the beginning, an initial meeting was held (November 2, 2022). Supporting information materials were disseminated, providing comprehensive details on the aim, scope, Delphi methodology, relevant terminology (notably the concept of PROMs), specified domains and subdomains, and information about the candidate PROMs derived from the SLRs, covering content validity and sensitivity to change, 3 aligned with COSMIN recommendations for evaluating PROMs 18 ( Supplementary material S1 , Chapters S1.1–S1.4 and S2.1–S2.3). During an anonymous online survey conducted via a REDCap interface 24 hosted by the University of Muenster, stakeholders voted twice (November 2022 to January 2023) and submitted comments in all rounds. This feedback was incorporated by the steering committee into subsequent Delphi rounds and consensus meetings. 14 , 21 Results from professionals and PWLE were analysed and presented separately. For further details on the Delphi process, see Supplementary material S1 , Chapters S3.2. and S3.3. PROMs endorsed by ≥50% of all panellists in the final Delphi round were included in a pre-final COS, which was presented, discussed, and finalised during the final online consensus meeting (March 7, 2023). A 50% threshold, adopted from our previous COS development and recommendations, 14 , 19 , 21 was chosen owing to the limited content validity of the found PROMs and the correspondingly expected lower acceptance.

Results from all conducted SLRs are detailed in Supplementary material S1 , Chapter S1.1.3 for pain intensity, including findings of previous SLRs, 25 , 26 Chapter S1.2.5 for physical function, Chapter S1.3.3 for self-efficacy, and Chapter S1.4.3 for adverse events. As recommended by COSMIN ( https://cosmin.nl/ ), 22 , 27 results of the evaluation of the quality of the PROM developmental process, using the COSMIN Risk of Bias Checklist ( Supplementary Box S1 ), are provided in Supplementary material S1 , Chapter S2.2, Tables S8–S13 . Further details on all PROMs, based on their developmental/inauguration articles, are provided in Supplementary material S1 , Chapter S2.3, Tables S14–S19 . In round 1, eight PROMs were evaluated: three generic PROMs (applicable across various procedures), four procedure-specific PROMs (designed for specific surgical contexts), and one generic PROM evaluated separately for each specific procedure. All PROMs were voted ‘appropriate’ (7–9 on a scale of 1–9) by >50% of both groups, professionals and those with lived experience ( Table 2 a). The same PROMs were reassessed in round 2, with all eight again voted ‘appropriate’ by >50% of both groups ( Table 2 a and Fig. 2 a). For the subdomain pain intensity in general, the PROMs ‘pain intensity on average’ and ‘worst pain intensity during the last 24 hours’ were rated ‘appropriate’ by 83.9% of voters. For the subdomain pain intensity at rest, the generic PROM ‘current pain intensity at rest’ was voted ‘appropriate’ by 87.1%, and for the subdomain pain intensity during activity, the procedure-specific PROMs were voted ‘appropriate’ by 96.8%. Pain intensity during physiotherapy was excluded, receiving <50% ‘appropriate’ votes for three of the four surgical procedures, with many comments in both rounds stating that ‘physiotherapy’ was not specific enough. Thus, three generic PROMs aligned with the respective pain intensity subdomains, and one procedure-specific PROM for each procedure proceeded to the final COS of PROMs ( Fig. 2 a, green-bordered box; Fig. 3 ). Table 2 Results of Delphi round voting. Overall voting results of both rounds (in % of panel members participating) for considering the specific PROMs as appropriate (7–9) for all panel members (All), for health professionals (HPs; i.e. HTA specialists, researchers, and clinicians) and those with lived experience (PWLE). Type of surgery: breast surgery (BS), sternotomy (ST), total knee arthroplasty (TKA), and endometriosis-related surgery (ES), all procedures (All). PROMs: physical functioning: BBUSQ-22, Birmingham Bowel and Urinary Symptoms Questionnaire; DASH, Disabilities of the Arm, Shoulder and Hand; FDQ, Functional Difficulties Questionnaire; OKS, Oxford Knee Score; WOMAC, Western Ontario and McMaster Universities Osteoarthritis Index. Self-efficacy: ASES, Arthritis Self-Efficacy Scale; ASES-8, Arthritis Self-Efficacy Scale- Short Form; K-SES, Knee Self-Efficacy Scale; RASE, Rheumatoid Arthritis Self-Efficacy Scale; ASES IMI-PainCare PROMPT adaptation. Adverse event: OR-SDS, Opioid Related Symptom Distress Scale. ∗POD1 (for POD3 and all later time points ‘since your surgery’ are replaced with ‘during the last 24 h’). Table 2 Fig 2 Final voting results of the second round for each domain: (a) pain intensity, (b) physical function, (c) self-efficacy, and (d) adverse events. Inclusion criteria for a PROM to be included in the final COS were that ≥50% of the panel rated the PROM between 7 and 9 (‘appropriate’, green). The vertical red dotted line indicates the 50% cut-off threshold for inclusion, and the PROMs ultimately included (after final discussion in the second meeting) are highlighted in the green-bordered box. Rating scale consists of 1–3, not appropriate (red); 4–7, potentially appropriate (blue); and 7–9, appropriate (green). all, all types of surgery; ASES-8, Arthritis Self-Efficacy Scale Short Form; ASES, Arthritis Self-Efficacy Scale; BBUSQ-22, Birmingham Bowel and Urinary Symptoms Questionnaire; BS, breast surgery; DASH, Disabilities of the Arm, Shoulder and Hand; ES, endometriosis related surgery; FDQ, Functional Difficulties Questionnaire; K-SES, Knee Self-Efficacy Scale; OKS, Oxford Knee Score; OR-SDS, Opioid-Related Symptom Distress Scale; PROM, patient-reported outcome measure; PROMPT, Providing Standardized Consented PROMs for Improving Pain Treatment; RASE, Rheumatoid Arthritis Self-Efficacy Scale; ST, sternotomy; TKA, total knee arthroplasty; WOMAC, Western Ontario and McMaster Universities Osteoarthritis Index. Fig 2 Fig 3 Final COS of PROMs enabling assessment of efficacy and effectiveness of pain management in the acute phase after surgery. BS, breast surgery; COS, core outcome set; ES, endometriosis-related surgery; PROM, patient-reported outcome measure; ST, sternotomy; TKA, total knee arthroplasty. Fig 3 Results of Delphi round voting. Overall voting results of both rounds (in % of panel members participating) for considering the specific PROMs as appropriate (7–9) for all panel members (All), for health professionals (HPs; i.e. HTA specialists, researchers, and clinicians) and those with lived experience (PWLE). Type of surgery: breast surgery (BS), sternotomy (ST), total knee arthroplasty (TKA), and endometriosis-related surgery (ES), all procedures (All). PROMs: physical functioning: BBUSQ-22, Birmingham Bowel and Urinary Symptoms Questionnaire; DASH, Disabilities of the Arm, Shoulder and Hand; FDQ, Functional Difficulties Questionnaire; OKS, Oxford Knee Score; WOMAC, Western Ontario and McMaster Universities Osteoarthritis Index. Self-efficacy: ASES, Arthritis Self-Efficacy Scale; ASES-8, Arthritis Self-Efficacy Scale- Short Form; K-SES, Knee Self-Efficacy Scale; RASE, Rheumatoid Arthritis Self-Efficacy Scale; ASES IMI-PainCare PROMPT adaptation. Adverse event: OR-SDS, Opioid Related Symptom Distress Scale. ∗POD1 (for POD3 and all later time points ‘since your surgery’ are replaced with ‘during the last 24 h’). Final voting results of the second round for each domain: (a) pain intensity, (b) physical function, (c) self-efficacy, and (d) adverse events. Inclusion criteria for a PROM to be included in the final COS were that ≥50% of the panel rated the PROM between 7 and 9 (‘appropriate’, green). The vertical red dotted line indicates the 50% cut-off threshold for inclusion, and the PROMs ultimately included (after final discussion in the second meeting) are highlighted in the green-bordered box. Rating scale consists of 1–3, not appropriate (red); 4–7, potentially appropriate (blue); and 7–9, appropriate (green). all, all types of surgery; ASES-8, Arthritis Self-Efficacy Scale Short Form; ASES, Arthritis Self-Efficacy Scale; BBUSQ-22, Birmingham Bowel and Urinary Symptoms Questionnaire; BS, breast surgery; DASH, Disabilities of the Arm, Shoulder and Hand; ES, endometriosis related surgery; FDQ, Functional Difficulties Questionnaire; K-SES, Knee Self-Efficacy Scale; OKS, Oxford Knee Score; OR-SDS, Opioid-Related Symptom Distress Scale; PROM, patient-reported outcome measure; PROMPT, Providing Standardized Consented PROMs for Improving Pain Treatment; RASE, Rheumatoid Arthritis Self-Efficacy Scale; ST, sternotomy; TKA, total knee arthroplasty; WOMAC, Western Ontario and McMaster Universities Osteoarthritis Index. Final COS of PROMs enabling assessment of efficacy and effectiveness of pain management in the acute phase after surgery. BS, breast surgery; COS, core outcome set; ES, endometriosis-related surgery; PROM, patient-reported outcome measure; ST, sternotomy; TKA, total knee arthroplasty. In round 1, nine procedure-specific PROMs were evaluated: two for breast surgery (BS), sternotomy (ST), and endometriosis-related surgery (ES), and three for total knee arthroplasty (TKA). In addition, two generic PROMs were evaluated separately for each procedure. All but four PROMs (OKS, BBUSQ, DASH, and FDQ) were rated ‘appropriate’ by >50% of both groups ( Table 2 b). Following comments and suggestions from round 1, an additional PROM (interference items from the BPI) was added in round 2 for each surgical procedure. In round 2, all PROMs except the BPI interference scale, the WOMAC and the same four voted out in the first round (OKS, BBUSQ, DASH, and FDQ) were again voted ‘appropriate’ by >50% of both groups ( Table 2 b). Consequently, two PROMs, one generic and one procedure-specific PROM for each surgery, were included in the final COS of PROMs ( Fig. 2 b, green-bordered box; Fig. 3 ). In round 1, all proposed PROMs were rated ‘appropriate’ by <50% of both groups for all four surgeries and by 50% from those with lived experience ( Table 2 c). Many participants commented that the existing PROMs assessing self-efficacy in people with acute postoperative pain were inadequate. Therefore, in round 2, participants were asked to recommend the PROM for self-efficacy ‘without restrictions’ or ‘on the condition that the PROM will be replaced immediately when a more suitable one is available’ (with restrictions) or to express their inability to decide. Sixty-six percent of all participants agreed to recommend the PROM with restrictions ( Table 2 c). The ASES IMI-PainCare PROMPT adaptation received the most ‘appropriate’ votes for all surgeries in round 2 (55.2% for TKA, BS, and ES and 48.3% for ST). Although it did not meet the 50% threshold for ST, the lower proportion of votes from those with lived experience was attributed to ‘unable to rate’ responses rather than outright rejection. Compared with all other items in this domain, which scored <20.7% ‘appropriate’ ( Table 2 c), the ASES IMI-PainCare PROMPT adaptation demonstrated relatively strong agreement and was therefore included in the COS for all types of surgery ( Fig. 2 c, green-bordered box; Fig. 3 ), contingent upon immediate replacement when a more suitable PROM becomes available (‘with restrictions’; Fig. 3 ). In round 1, only the OR-SDS IMI-PainCare PROMPT adaptation for TKA and the OR-SDS original scale for ES were voted ‘appropriate’ by >50% of both groups ( Table 2 d). In addition, for the adverse events domain, participants were asked about appropriateness of including PROMs in this COS; 64.5% (62.1% HP, 100% PWLE) chose ‘The IMI-PainCare PROMPT COS assessment tool should contain always both PROMs and Clinician-Reported Outcomes’ ( Table 2 d). In round 2, as with the domain self-efficacy, participants were asked whether they would recommend the PROMs without restrictions or only on the condition of replacement whenever possible or whether they were undecided before evaluating them. Fifty-eight percent of all participants agreed to recommend the PROMs only on the condition of replacement whenever possible ( Table 2 d). Again, there was a notably high share of participants selecting ‘unable to rate’ in round 2, particularly among those with lived experience (66.67%; 15.4–30.8% of HP; Table 2 d). In round 2, only the OR-SDS IMI-PainCare PROMPT adaptation was voted ‘appropriate’ by >50% of the entire group for ES (51.7%). For all other types of surgery, this PROM narrowly missed the threshold (48.3% for TKA and ST and 44.8% for BS), but it was rated ‘appropriate’ by 50% of professionals for TKA and ST. As with self-efficacy, the lower ratings from those with lived experience were attributed to ‘unable to rate’ responses rather than rejection. Thus, the OR-SDS IMI-PainCare PROMPT adaptation for all types of surgery was included in the COS ( Fig. 2 d, green-bordered box; Fig. 3 ), also contingent upon immediate replacement when a more suitable PROM becomes available (‘with restrictions’; Fig. 3 ). Figure 3 presents the final COS of PROMs for postoperative assessment across the four outcome domains, including subdomains related to pain intensity, as discussed and consented during the final consensus meeting. Table 3 provides the precise phrasing of the PROMs, including the descriptors, ranges, and anchors. The complete questionnaires with all PROMs of the COS are available in Supplementary material S2 . Table 3 IMI-PainCare PROMPT Final Core Outcome Set of PROMs (exact wording of questions, range of scales, anchors, and descriptors). BS, breast surgery; ES, endometriosis-related surgery; PROM, patient-reported outcome measure; ST, sternotomy; TKA, total knee arthroplasty. Table 3 IMI-PainCare PROMPT Final Core Outcome Set of PROMs (exact wording of questions, range of scales, anchors, and descriptors). BS, breast surgery; ES, endometriosis-related surgery; PROM, patient-reported outcome measure; ST, sternotomy; TKA, total knee arthroplasty.

Led and conceptualised the project: EMP-Z Contributed to the conceptualisation of the project: HL, WM, UK Developed the protocols for the consensus process: EMP-Z Developed the protocols for the systematic reviews: UK, EMP-Z Performed the assessments of psychometric properties of the PROMs: EMP-Z, SDL, DR, PZ, UK, HH Planned and coordinated the consensus process, including the consensus meetings and Delphi process: EMP-Z Contributed to development of the consensus meeting: SDL, LH, KV, JV, PZ, UK, WM, HL, DR, CW Contributed to the organisation of the consensus meetings and the Delphi process: CW, HL, DR, UK Analysed the results of the Delphi process: EMP-Z, DR Analysed the PROMs related to the sensitivity-to-change properties: JV Drafted the first version of the manuscript: EMP-Z, DR, SDL Coordinated all follow-up versions, including the final version: EMP-Z Was involved in the follow-up versions, including the final version: SDL, DR Revised the manuscript: LH, HL, WM, KV, JV, PZ, UK, CW Read and commented on the final version of the manuscript: all authors

This work has been conducted as part of IMI-PainCare PROMPT (an Innovative Medicines Initiative 2 Joint undertaking under grant agreement No 777500). This Joint Undertaking receives support from the European Union's Horizon 2020 research and innovation programme and EFPIA. www.imi.europa.eu ; www.imi-paincare.eu . The statements and opinions presented here reflect the author’s view and neither IMI nor the European Union, EFPIA, or any Associated Partners are responsible for any use that may be made of the information contained therein.

An international, multiprofessional stakeholder panel, including individuals with lived experience, developed consensus on a COS of PROMs for assessing the effectiveness and efficacy of perioperative pain management in adults. For each of the four predefined key domains, reflecting the biopsychosocial complexity of pain management, 21 at least one PROM was identified for use across all surgical procedures ( Fig. 3 , left side; Table 3 ). In addition, procedure-specific PROMs were recommended for two domains: pain intensity during activity and physical function ( Fig. 3 , right side; Table 3 ). For pain intensity during activity, one PROM was selected for each exemplar surgical procedure, aligning with recent requests to assess pain intensity not only at rest but also during activity using ‘common, procedure-specific pain-evoking manoeuvres’. 8 Similarly, for physical function, the COS of PROMs includes easy-to-use single-item tools, with one PROM for all surgical procedures and one specific to each surgical procedure. For procedures not covered, the general PROMs can be used, although further procedure-specific PROMs require development in the future. The PROMs presented to the panel during the consensus process underwent systematic identification and rigorous evaluation, including content validity and examination of their development process. 18 , 19 , 20 Several PROMs included in the voting process demonstrated a significant lack of content validity, raising concerns about whether they accurately measure the intended constructs. 28 This challenge is common in COS development, including those addressing pain. 29 To ensure responsiveness of the PROMs, we evaluated adaptations and face-valid scales for sensitivity to change. 3 Nevertheless, the included PROMs require further evaluation or replacement with higher-quality PROMs when available, as recommended by this Delphi panel for the PROMs selected for the domains adverse events and self-efficacy (‘with restrictions’; Fig. 3 ). COS updates are a common part of the development process, enabling re-evaluation and refinement over time. 19 The four exemplar surgical procedures reflect a broad spectrum of surgical trauma and conditions, patient demographics, pain mechanisms, and treatment approaches. 3 , 21 We recommend adopting this COS in research and clinical settings while acknowledging flexibility to include additional outcomes as required for specific surgeries or health conditions. Beyond acute postoperative pain, COS have been developed for other painful conditions, such as complex regional pain syndrome, 28 low back pain, 29 , 30 and perioperative pain in trials on regional anaesthesia. 31 Overarching pain-related COS of domains, such as developed recently by the INTEGRATE group for acute, acute-to-chronic, chronic, and recurrent pain, 32 or by IMMPACT for chronic pain in general, 33 facilitate data harmonisation and comparability across different pain conditions. These COS reflect several aspects of the biopsychosocial complexity of both acute and chronic pain by additionally integrating psychological domains, such as self-efficacy, alongside physical function or quality of life. This better reflects patient experiences of pain, as raised by those with lived experiences during our initial domain process. 21 Recent studies have shown that relying solely on pain intensity as an outcome measure can lead to unidimensional pain management, including overprescription of opioids. 5 , 6 , 7 , 31 Integrating multiple domains beyond pain intensity will guide future pain management towards comprehensive, targeted and multimodal approaches rather than reliance on opioid use. Harmonising PROMs across domains simplifies their assessment by guiding the selection of measurement tools. However, guidance for treatment through implementation of the COS of PROMs in clinical practice is complex and will require defining treatment thresholds and minimal clinically important differences for each PROM. Certain limitations apply to our COS development process. Although the distribution of panel members was uneven across locations and countries of origin, the professional and personal diversity of all members, and inclusion of those with lived experiences, ensured a broad range of perspectives, reflected in lively and comprehensive discussions and leading to a balanced consensus despite the relatively small panel size. Following available recommendations, 18 , 19 we carefully ensured panel diversity by including representatives from various stakeholder groups (researchers, clinicians, HTA experts, and persons with lived experience) while limiting the total to 40 participants to balance broad perspectives with practical feasibility for consensus meetings. Consistency in panel composition throughout the COS development process 2 , 21 allowed for knowledge transfer, and the strong engagement of those with lived experience, whose input was prioritised in cases of divergent voting, helped to address the limitation of their smaller numbers. The flexible and multi-round nature of the Delphi process poses risks of participant dropout and bias towards majority views, particularly in smaller panels. 34 To mitigate this, we provided extensive background information on all PROMs and the process and gave feedback to panellists after each round to enable informed decision-making. Furthermore, we considered and discussed participants' reluctance to rate certain PROMs and near-threshold voting results when finalising the COS. In addition, votes from professionals and those with lived experience were presented separately to the panel. Cross-cultural translation ensures items perform well after translation or cultural adaptation, which is relevant for global usability of the COS. 27 Although our COS and the development process of the included PROMs do not fully address patient diversity, such as the full range of socioeconomic status, education, race, or ethnicity, we tested all PROMs for sensitivity to change in a prospective study 3 and translated them into French, German, Italian, Serbian, Finnish, Spanish, and Portuguese 35 to facilitate cross-cultural adaptation studies. The IMI-PainCare PROMPT core outcome sets of patient-reported outcome measures for assessing the effectiveness and efficacy of perioperative pain management was completed. Despite some restrictions, it provides a set of outcome measures enabling a comprehensive and harmonised assessment of postoperative pain, extending beyond common pain intensity scales. Simple, easily applicable tools such as single-item scales for postoperative pain intensity and pain-related physical function were considered most appropriate. Adopting this COS will align outcome assessment in acute perioperative pain, thus enhancing comparability and generalisability of clinical trial results, advancing translation of results from clinical practice to research and back, reducing reporting bias, and ensuring relevance of outcome assessment to all stakeholders, including patients.

During the preparation of this work, the authors used the local large language model of the University of Muenster Llama 3–70B to check orthography, punctuation, and grammar. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.

EMP-Z has received payments from Grünenthal (Germany), MSD Sharp & Dohme GmbH (Germany), and Medtronic (UK) for advisory board activities and lecture fees. EMP-Z has received funding for research from the Gemeinsamer Bundesausschuß (GBA), the Innovative Medicines Initiative 2 Joint Undertaking under grant agreement 777500 (this Joint Undertaking receives support from the European Union's Horizon 2020 research and innovation programme and EFPIA), Grünenthal Germany, the DFG, and the BMBF. All payments were made to the institution (UKM), not to herself. HL was an employee of Gruenenthal GmbH, Aachen, Germany, at the time of the project. WM has received payments for lectures from Kyowa, Mundipharma, Grünenthal and Spectrum Therapeutics (total funds ∼€4000). WM has received payments for consultancy from Merck, Sanofi, MSD, Tafalgie, Mundipharma, Grünenthal, and Ethypharm. WM's institution received funding for research from European Commission, Gemeinsamer Bundesausschuß (GBA), Medtronic, Pfizer, Mundipharma, Grünenthal, and Vertanical. KV has received payments to her institution for consultancy, lectures, and associated travel expenses from AbbVie, Bayer, Eli Lilly, Gedeon Richter, and Reckitts. JV is conducting research sponsored by Viatris and AstraZeneca. The other authors declare that they have no conflicts of interest.