Rates of discontinuation and non-publication of regional anaesthesia clinical trials: a registry-based cross-sectional analysis

Abstract

Ultrasound-guided peripheral nerve, spinal and epidural anaesthesia deliver site-specific analgesia that lowers opioid use and the incidence of chronic postsurgical pain. Portable high-resolution ultrasound, introduced in the mid-2000s, enables direct needle visualisation and broadened safe block options.

Regional anaesthesia research publications have risen approximately 20-fold since the 1980s, with the steepest growth occurring after 2016 [1]. Randomised controlled trials, however, are often discontinued prematurely or remain unpublished. A 2022 meta-analysis of 326 protocols showed 30% stopped early and 21% were unpublished after 10 years, potentially wasting resources and slowing progress [2]. Anaesthesia mirrors this pattern; of 1052 abstracts presented at American Society of Anesthesiologists meetings (2001–2004), only 54% became papers and ‘positive’ trials were 42% more likely to publish [3]. Fewer than one in six completed anaesthesia trials on ClinicalTrials.gov post results, and almost half of published trials contain major outcome or sample-size discrepancies [4, 5]. Selective reporting skews meta-analyses, slows uptake of new blocks and undermines participant altruism. Therefore, we analysed registered regional anaesthesia studies to quantify completion and publication rates and identify factors linked to early termination or non-publication. Illuminating where and why trials go missing may guide practices that protect participant trust and strengthen the regional anaesthesia evidence base.

We searched ClinicalTrials.gov for phase 3–4 randomised controlled trials of regional anaesthesia. The strategy combined Medical Subject Heading (MeSH) and free-text terms for peripheral nerve, neuraxial and ultrasound-guided blocks (online Supporting Information Appendix S1). Trials with primary completion before 10 March 2023 (thereby guaranteeing ≥ 24 months for potential publication) and status ‘Completed’; ‘Terminated’; ‘Suspended’; ‘Withdrawn’; or ‘Unknown’ were eligible. For each trial we examined the registry ‘Publications’ field. If no matching citation appeared, three reviewers (TH, EB, AT) searched MEDLINE (via PubMed), Embase and Google Scholar independently using the title and the ClinicalTrials.gov identifier (NCT number). Publication was confirmed when a full manuscript or abstract, retracted papers included, matched the protocol. If a publication was absent, we emailed the study contact, principal investigator, chair and co-ordinator (sequentially) listed on ClinicalTrials.gov (weekly for 3 weeks, template in online Supporting Information Appendix S2). No response after 6 weeks classified ‘Terminated’/‘Suspended’/‘Withdrawn’ trials as ‘discontinued’ and ‘Completed’/‘Unknown’ trials as ‘not published’. If contact details were unavailable, the same rules applied. Raw data, correspondence and analyses are posted on Open Science Framework [6].

The search yielded 1254 registry entries; 790 met inclusion (Fig. 1). Of these, 561 (71%) were ‘Completed’; 129 (16%) ‘Unknown’; 60 (8%) ‘Terminated’; and 40 (5%) ‘Withdrawn’. Most trials were international (521/790, 66%); 195 (25%) were US-based; 70 (9%) lacked a location; and 4 (0.5%) spanned US and non-US sites. Median (IQR [range]) enrolment totalled 76,478 (80 (50–120 [0–2108])) participants. Peer-reviewed publications existed for 501 trials (63%), leaving 289 (37%) unpublished. Publication success by ClinicalTrials.gov status was as follows: ‘Completed’ 416/561 (74%); ‘Unknown’ 56/129 (43%); ‘Terminated’ 21/60 (35%); and ‘Withdrawn’ 8/40 (20%). Unpublished trials involved 25,041 participants and listed 1062 adverse events. We contacted 308 trialists: 52 lacked a retrievable email address and 280 did not reply. Among 28 responders, the main barriers were limited time and funding (Table 1). Seven provided publications that were subsequently verified and added.

Regional anaesthesia trials showed a comparatively strong publication record. This rate (63.4%) exceeds the 29% publication rate observed in colorectal cancer randomised controlled trials tracked with the same method and dwarfs historical anaesthesia conference data, showing that positive abstracts were twice as likely to publish as neutral ones [7]. Yet one in three studies, representing > 25,000 participants, and 16% of registry records with ‘Unknown’ status still fail to inform practice, underscoring persistent ethical and scientific waste.

Selective survival further narrows the evidence base. Only 35% of ‘Terminated’ and 20% of ‘Withdrawn’ trials were published, suggesting that null or under-powered studies may face steeper editorial hurdles. Trialists cited limited time and funding most frequently, suggesting that modest incentives, waived article processing charges for negative or inconclusive studies, mandatory registry-linked manuscripts and credible enforcement of the US Food and Drug Administration Amendments Act (FDAAA) civil penalty provisions, could accelerate dissemination. Article processing fees have been documented as a significant barrier to publication, particularly for investigators with limited resources. Compliance audits likewise show that reporting rates improved after the section 801 of the Food and Drug Administration Amendments Act (2017) clarified monetary penalties for non-compliance [8]. Further waste can be curtailed by adopting adaptive platform designs that recycle control groups and preserve data from discontinued groups without compromising validity. Ultimately, achieving complete transparency and ensuring the findings of all trials are publicly accessible, will honour participant altruism, reduce avoidable research waste and secure a trustworthy evidence base for regional anaesthesia practice.