Awareness and acceptability of Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials core outcome set for chronic pain among surveyed neuropathic pain authors

This study was the first to assess the awareness and acceptability of the IMMPACT-recommended core set of outcome domains and COMs for chronic pain among authors of SRs and RCTs about interventions for management of NeuP. The knowledge of existing COS for chronic pain among surveyed authors was suboptimal; 73% of surveyed SR authors and 56% of trialists were familiar with the IMMPACT core set of domains, which was first developed in 2003 [3], and around three-quarters of respondents could also correctly identify the number of domains it contains. All RCT authors correctly identified pain, and almost all SR authors identified pain, physical functioning and emotional functioning as domains that belong to the IMMPACT COS. This shows that the majority of respondents are aware of the existence of the COS and domains it contains to some extent, and indicates at other problems preventing consistent and complete use of IMMPACT-recommended COS for chronic pain among NeuP authors.

Importantly, SR authors agreed that five of six IMMPACT-proposed outcome domains should be included in the core set (except participant disposition), as well as one additional outcome domain (role functioning), which is currently not a part of IMMPACT COS. Although such consensus could not be reached among RCT authors, the majority of respondents, represented by SR authors, acknowledge the critical importance of IMMPACT-proposed core outcome domains for chronic pain.

The knowledge of the core set of outcome measures, first published in 2005 [4], differed among SR and RCT authors. Less than half of trialists were familiar with IMMPACT-recommended COMs for chronic pain and consensus for inclusion of any of the listed outcome measures in a core set could not be reached among RCT authors. On the contrary, the majority of SR authors were aware of the IMMPACT-recommended COMs and almost all could identify PGIC as one of the core measures. Consensus among SR authors for inclusion in the core set was reached for two IMMPACT-recommended COMs (PGIC and 11-point NRS), and one non-IMMPACT outcome measure (EQ-5D). This shows that surveyed SR authors feel that EQ-5D is a critically important outcome measure and should be considered for inclusion in a core set of outcome measures.

In the original manuscripts describing analyzed COS and COMs, it was not reported how consensus needed to include outcome domains and measures in the core set during IMMPACT consensus meetings was reached [3,4]. Therefore, we do not know the exact proportion of specialists that voted for inclusion of each outcome domain and measure into a COS. Although the choice of thresholds is somewhat subjective, it is generally accepted that if 70% or more of the respondents score an outcome as critically important and <15% as of limited importance, it should be included in a COS [2], which was the criterion we used as well.

Despite existence of relevant COS and COMs, our results point out at inconsistent adoption of IMMPACT recommendations among surveyed authors, as neither SR nor RCT authors used the full COS and COMs while choosing outcome domains and measures in their studies, that is, when used, the majority used it only partially.

The lack of awareness of all COS domains was stated as the main perceived reason, which prevented the consistent use of all COS domains recommended by the IMMPACT among both groups of respondents. The main obstacle in consistent use of IMMPACT-recommended COMs among trialists was also the lack of awareness, and for SR authors it was the lack of proposed COMs in included RCTs. Both SR and RCT authors stated that the incentive for using recommended COS and COMs when designing future studies would be facilitated conduct of systematic reviews/meta-analyses due to homogeneous outcome domains and measures.

It is important to periodically review the proposed COS to ensure that agreed outcomes are still relevant, to supplement them with the new outcomes as needed, to check whether the process of implementation has been successful and to engage further stakeholders [2]. Our study can be used as a part of this periodic review of the proposed COS and COMs, and it identified an additional outcome domain and additional outcome measure for which respondents feel are critical for inclusion in the core sets.

In the published literature there are very few studies that have investigated the use and uptake of any COS. Kirkham et al. conducted a survey of the use of rheumatoid arthritis (RA) COS published after its introduction in 1994 among RA trialists. They showed the increase in the number of trials reporting on the COS over time [8,9], but also that some trialists did not measure the whole COS despite being aware of it [8]. It was postulated that the increased adherence to RA COS was encouraged by the introduction of regulatory guidance from the agencies such as the US FDA [10] and EMA [11] involved in ratifying the COS in that field. Perhaps a similar approach could be used in the field of chronic pain.

Another study by Kirkham et al. in 2017 [9] used the new method to analyze the uptake of the RA COS and recommended the use of the clinical trial registry information for assessing uptake of the COS because it was more reliable than citation analysis, and more efficient and up-to-date than assessing publications of the registered trials.

The implementation of Outcome Measures in Rheumatology (OMERACT) COS in hip and/or knee osteoarthritis trials [12,13] and gout [14] trials has also been reported as limited. Another example of low uptake of COS comes from trials on fall injury prevention in older people. Copsey et al. [15] appraised the use of The Prevention of Falls Network Europe (ProFaNE) COS for trials on fall injury prevention in older people and found only 1 of 34 studies that reported use of the full ProFaNE COS, as well as inconsistency in the outcome measures used.

A recent study surveyed the knowledge and appropriateness of the Pediatric Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials (PedIMMPACT)-recommended core outcome domains for acute pediatric pain among SRs and RCTs authors. The findings indicated that the majority of participating authors were not familiar with the six domains of PedIMMPACT COS for acute pediatric pain, did not use the COS while preparing their studies and gave preference to outcome domains, which do not completely overlap with PedIMMPACT COS [16]. Our results, based on the sample of respondents that published RCTs and SRs in the field of NeuP, show that although the authors are aware of the IMMPACT COS, its implementation has not been successful.

There are possible actions that should be taken in order to improve the use of COS among authors of RCTs and SRs. As proposed recently, researchers would more likely consider using COS if its use would impact the likelihood of getting their research funded and published, or have an effect on clinical policy or reimbursement [1]. That identifies several stakeholders who could be targeted to improve implementation of existing COS; such as research funders, journal editors, clinical guideline developers, health technology assessment organizations and payers [1]. An example of action of research funders can be found in guidance for applicants to the UK’s National Institute for Health Technology Assessment [17], which states: ‘where established Core Outcomes exist they should be included amongst the list of outcomes unless there is good reason to do otherwise. Please see The COMET Initiative website [18] to identify whether Core Outcomes have been established.’

Another way to increase the uptake is to require a statement of COS adherence while registering protocols of RCTs and SRs, and justification of exclusion of certain COS domains, if applicable. For example, the SPIRIT Statement (Standard Protocol Items: Recommendations for Interventional Trials) provides guidance for reporting protocols of clinical trials [19] in which, among other relevant items, trialists are encouraged to determine if a relevant COS exists, and if so, to include proposed outcomes in their trial.

As hypothesized previously, the introduction of guidance from regulatory agencies, which were involved in defining the COS for RA, such as the FDA and EMA, might have contributed to increased uptake of core outcomes among trialists in RA [8]. Such approach can be applied to other COS as well. A support for that approach stems from our finding that SR authors stated the lack of regulatory guidance as the third most common reason for not using the proposed COS.

As pointed out in a recent editorial, developers of COS should proceed with advanced stages of COS development to ensure that proposed COS is still relevant, with the ultimate goal of ensuring comparability of future trials and systematic reviews [6].

Patient and caregivers’ involvement in the development of COS is especially relevant for comparative effectiveness research, where long-term patient-reported outcomes are the important end points, and has been noted in recent years [20]. Although patients represent key stakeholders, they were not included in the IMMPACT COS development. This might have affected participants’ decision to use the IMMPACT COS, even though none of the participants indicated so when asked about reasons for inconsistent use of the IMMPACT COS.

Although not within the scope of this study, investigating participants’ views of the COS development methodology would be the next step, which should be addressed in the process of investigating reasons for insufficient uptake of the proposed COS.

Our results are based on a limited number of samples of SR and RCT authors; particularly low response rate was obtained from authors of RCTs. That is not surprising, considering that the cohort of RCTs whose authors we contacted were conducted much earlier than SRs, and therefore there is lower chance of successfully achieving contact with the authors. Furthermore, it is possible that our results are not generalizable to knowledge and opinions of all SR and RCT authors from this field, and that there was selection bias, with self-selected participants who might have been more motivated to participate in the study, with more knowledge about the field. Still, these results indicate that this bias can be considered minimal; as not all participants exhibited adequate knowledge about the questions they were served. Since we invited an entire group of SR authors to participate, it is possible that we obtained more than one response from the same group of SR authors. It would be interesting to analyze the impact on results, but we could not collect these data as we guaranteed anonymity to the survey participants.

Another limitation is provisional division of participants to authors of SRs and RCTs based on first identification of SR authors and subsequent exclusion of the same participants if they appeared to be on the list of RCT authors as well. That might have skewed the number of respondents in favor of SR authors.

The implementation of IMMPACT-recommended COS and COMs for chronic pain among surveyed authors of RCTs and SRs in the field of NeuP has not been successful. Neither SR nor RCT authors used the full COS and COMs while choosing outcome domains and measures in their studies, that is, when used, the majority used them only partially. If a relevant COS is not widely implemented, its benefits for comparative effectiveness research are not realized since comparability and generalizability of research findings are limited. Therefore, actions should be taken in order to improve the use of COS among authors of RCTs and SRs in the field of NeuP, or to reconsider their adequacy.

We are grateful to anonymous participants who invested their time to respond to survey.

This study was funded by the Croatian Science Foundation, grant no. IP-2014-09-7672 ‘Professionalism in Health Care’, awarded to Ana Marušić. The funder had no role in the design of this study, during its execution and data interpretation. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.

The authors state that they have obtained appropriate institutional review board approval or have followed the principles outlined in the Declaration of Helsinki for all human or animal experimental investigations. In addition, informed consent has been obtained from the participants involved and the participants could not access the survey without providing their informed consent.