Crowd-sourcing and automation facilitated the identification and classification of randomized controlled trials in a living review

OBJECTIVES

To evaluate an approach using automation and crowdsourcing to identify and classify randomized controlled trials (RCTs) for rheumatoid arthritis (RA) in a living systematic review (LSR).

METHODS

Records from a database search for RCTs in RA were screened first by machine learning and Cochrane Crowd to exclude non-RCTs, then by trainee reviewers using a Population, Intervention, Comparison, and Outcome (PICO) annotator platform to assess eligibility and classify the trial to the appropriate review. Disagreements were resolved by experts using a custom online tool. We evaluated the efficiency gains, sensitivity, accuracy, and interrater agreement (kappa scores) between reviewers.

RESULTS

From 42,452 records, machine learning and Cochrane Crowd excluded 28,777 (68%), trainee reviewers excluded 4,529 (11%), and experts excluded 7,200 (17%). The 1,946 records eligible for our LSR represented 220 RCTs and included 148/149 (99.3%) of known eligible trials from prior reviews. Although excluded from our LSRs, 6,420 records were classified as other RCTs in RA to inform future reviews. False negative rates among trainees were highest for the RCT domain (12%), although only 1.1% of these were for the primary record. Kappa scores for two reviewers ranged from moderate to substantial agreement (0.40-0.69).

CONCLUSION

A screening approach combining machine learning, crowdsourcing, and trainee participation substantially reduced the screening burden for expert reviewers and was highly sensitive.

Prioritization of clinical questions for the Australian Living Guideline for the Pharmacological Management of Inflammatory Arthritis

AIM

Living guidelines aim to reduce delays in translating new knowledge into practice by updating individual recommendations as soon as relevant new evidence emerges. We surveyed members of the Australian Rheumatology Association (ARA) to develop a list of priority questions for the Australian Living Guideline for the Pharmacological Management of Inflammatory Arthritis (ALG) and to explore clinicians' use of clinical practice guidelines.

METHODS

An electronic survey of ARA members was performed in two phases. The first survey contained questions about current guideline use and beliefs and invited participants to submit at least three questions relevant to the management of rheumatoid arthritis (RA). In the second round, participants selected 10 questions they considered to be the highest priority from the collated list and ranked them in priority order. The sum of ranks was used to generate a final priority list.

RESULTS

There were 115 (21%) and 78 (14%) responses to the first and second survey rounds respectively. 87% of respondents use existing rheumatology guidelines in their usual practice, primarily EULAR guidelines. Most respondents favored the development of Australian rheumatology guidelines. In total, 34 potential recommendation topics were identified and ranked in order of priority.

CONCLUSION

A list of 34 clinical questions about RA management, ranked in order of importance by clinicians, has informed the development of the ALG. Similar prioritization exercises in other contexts may permit guidelines to be tailored to the needs of guideline users in their specific context, which may facilitate international collaboration and promote efficient translation of evidence to practice.

Australian recommendations on perioperative use of disease-modifying anti-rheumatic drugs in people with inflammatory arthritis undergoing elective surgery

Disease-modifying anti-rheumatic drugs (DMARDs) are effective treatments for inflammatory arthritis but carry an increased risk of infection. For patients undergoing surgery, there is a need to consider the trade-off between a theoretical increased risk of infection with continuation of DMARDs perioperatively versus an increased risk of disease flare if they are temporarily withheld. We used the Grading of Recommendations Assessment, Development and Evaluation methodology to develop recommendations for perioperative use of DMARDs for people with inflammatory arthritis undergoing elective surgery. The recommendations form part of the National Health and Medical Research Council-endorsed Australian Living Guideline for the Pharmacological Management of Inflammatory Arthritis. Conditional recommendations were made against routinely discontinuing conventional synthetic and biologic (b) DMARDs in the perioperative period but to consider temporary discontinuation of bDMARDs in individuals with a high risk of infection or where the impact of infection would be severe. A conditional recommendation was made in favour of temporary discontinuation of targeted synthetic DMARDs in the perioperative period.

Australian recommendations on tapering of biologic and targeted synthetic disease-modifying anti-rheumatic drugs in inflammatory arthritis

Biological and targeted synthetic disease-modifying antirheumatic drugs (b/tsDMARDs) have been an important advance in the management of inflammatory arthritis, but are expensive medications, carry a risk of infection and other adverse effects, and are often perceived as a burden by patients. We used GRADE methodology to develop recommendations for dose reduction and discontinuation of b/tsDMARDs in people with rheumatoid arthritis (RA), axial spondyloarthritis (AxSpA) and psoriatic arthritis (PsA) who have achieved a low disease activity state or remission. The recommendations form part of the Australian Living Guideline for the Pharmacological Management of Inflammatory Arthritis, an NHMRC-endorsed 'living' guideline, in which recommendations are updated in near real-time as new evidence emerges. Conditional recommendations were made in favour of dose reduction in RA and AxSpA but not in PsA. Abrupt discontinuation of b/tsDMARDs is not recommended in any of the three diseases. This article is protected by copyright. All rights reserved.

Experimental infections, using a fluorescent marker, of two elasmobranch species by unciliated larvae of Branchotenthes octohamatus (Monogenea: Hexabothriidae): invasion route, host specificity and post-larval development

The infection biology of Branchotenthes octohamatus (Monogenea: Hexabothriidae) from the gills of the southern fiddler ray, Trygonorrhina fasciata (Rhinobatidae), was studied using the fluorescent dye, 5(6)-carboxyfluorescein diacetate N-succinimidyl ester (CFSE). This is the first use of this technique on a monogenean species with unciliated larvae and the first for any monogenean larva infecting an elasmobranch host. CFSE-labelled post-larvae were recovered from gills of T. fasciata within 30 min of exposure to the host, providing strong evidence that larvae invade host gills directly and do not migrate after initial attachment elsewhere. The rapidity with which larvae settled suggests that the mode of infection may deliver larvae directly to the gills via the host's inhalant respiratory current. The specificity of B. octohamatus was investigated by exposing a sympatric rhinobatid host species, the western shovelnose ray, Aptychotrema vincentiana, to B. octohamatus larvae newly emerged from eggs laid by adult parasites from gills of T. fasciata. Experimental exposure of A. vincentiana to freshly hatched B. octohamatus larvae resulted in a persistent infection, indicating that B. octohamatus may not be strictly host specific. Post-larval development charted on these experimentally infected A. vincentiana specimens was slow. Parasites appeared to be sexually mature at 91 days at 21–25°C. Branchotenthes octohamatus larvae bear only 4 pairs of hooklets on the haptor whereas all other hexabothriid larvae described so far have 5 hooklet pairs. Ontogenetic changes to the haptor revealed that it is probably hooklet pair III that is lost from B. octohamatus prior to larval development.

A redescription of Calicotyle australis Johnston, 1934 (Monogenea: Monocotylidae) from the type-host Trygonorrhina fasciata (Rhinobatidae) off Adelaide, South Australia, including descriptions of live and silver stained larvae

Calicotyle australis Johnston, 1934 (Monogenea Monxocotylidae) is redescribed from the cloaca of the type-host, the southern fiddler ray Trygonorrhina fasciata (Rhinobatidae) off Adelaide, South Australia. Lobed glands joining the oötype are reported for the first time and may be characteristic of the genus. The presence of an appendix associated with the seminal vesicle in C. australis, previously reported as absent, is confirmed. The anatomy of the oncomiracidium of C. australis is described from observations of live larvae, and the number and distribution of ciliated epidermal cells and sensilla, revealed by silver staining larvae, is also described. Use of larval characters to distinguish between species of Calicotyle Diesing, 1850 and other closely related monocotylids is discussed.

Three unrelated species, 3 sites, same host – monogenean parasites of the southern fiddler ray, Trygonorrhina fasciata, in South Australia: egg hatching strategies and larval behaviour

The southern fiddler ray, Trygonorrhina fasciata (Rhinobatidae), is parasitized by 3 monogenean (platyhelminth) species from 3 families on 3 different sites of the host: Calicotyle australis (Monocotylidae) from the cloaca, Pseudoleptobothrium aptychotremae (Microbothriidae) from the skin and Branchotenthes octohamatus (Hexabothriidae) from the gills. Cues that promote egg hatching were investigated for each species and the behaviour of their larvae was also documented. Eggs were laid by parasites in vivo and maintained at 22 degrees C. Three different egg hatching and host finding strategies were discovered. Calicotyle australis eggs hatched spontaneously with a strong diurnal rhythm that is likely to be under circadian control. The larva is ciliated, photo-responsive and can survive for up to 24 h at 22 degrees C after hatching. Pseudoleptobothrium aptychotremae may have a 'bet-hedging' strategy. Some eggs hatched spontaneously and rhythmically. However, since the hatching success was low, it is possible that other eggs require a different cue provided by the host. The larva is also ciliated but shows no photo-response and was observed to remain active for <4 h at 22 degrees C after hatching. Branchotenthes octohamatus has a 'sit-and-wait' strategy that depends on mechanical disturbance to stimulate hatching. The larva is unciliated, shows no photo-response but may survive for more than 2 days at 22 degrees C after hatching. The implications of hatching strategy, larval behaviour and morphology in the goal to find a host are discussed for each species.

Branchotenthes octohamatus sp. n. (Monogenea: Hexabothriidae) from the gills of the southern fiddler ray, Trygonorrhina fasciata (Rhinobatidae) in South Australia: description of adult and larva

Branchotenthes octohamatus sp. n. (Monogenea: Hexabothriidae) is described from the gills of the southern fiddler ray, Trygonorrhina fasciata Müller et Henle (Elasmobranchii: Rhinobatidae), off Adelaide, South Australia. It is distinguished from the type species, Branchotenthes robinoverstreeti Bullard et Dippenaar, 2003, by producing eggs that are joined end to end forming a chain, in the morphology of the male copulatory organ that has a pronounced constriction in duct diameter between proximal and distal regions, the possession of a thin muscular layer surrounding the proximal part of the male copulatory organ and distal region of the vaginae, and by the absence of a raised process on the shaft of the hamulus. An amended generic diagnosis is provided and the reliability of sperm duct number as a generic character is discussed. The oncomiracidium of B. octohamatus is also described and is the first monogenean to be described with only eight hooklets in the larval haptor. This discovery of eight hooklets may be important for higher-level monogenean evolutionary hypotheses.