# Workload, Burnout and Safety in Clinical Teams: Four Reviews Read Against One Trial

> Four reviews on clinician burnout, AI in nursing, nursing quality and safety, and the WHO surgical checklist share a concern with workload but differ in design and in how much their evidence can support.

AMDana Desk · 2026-10-11 · https://blog.amdana.app/workload-burnout-and-safety-in-clinical-teams-four-reviews-read-against-one-tria

Four of the sources reviewed here examine how health care organizations change the conditions of clinical work. They cover interventions for clinician burnout [2], artificial intelligence (AI) in nursing [3], quality and safety recommendations for nursing practice [4], and the World Health Organization Surgical Safety Checklist (WHO SSC) [5]. Read together, they repeatedly return to workload, but the strength and reporting of their underlying evidence differ [2, 3, 4, 5]. A fifth source, a phase 3 randomized trial of the cardiac drug aficamten, addresses an unrelated clinical question. It is used here only as a reference point for what a prespecified randomized design reports [1].

## The five sources and their designs

The designs differ more than the topics do, and those differences affect how much weight each finding can carry [1, 2, 3, 4, 5].

- **Burnout and well-being interventions.** This is a systematic review, meaning a structured search and appraisal of published studies against stated eligibility criteria [2]. It covered studies since 2015 of physicians, nurses and allied healthcare professionals that reported quantifiable preintervention and postintervention outcomes using validated well-being measures [2]. The authors searched Medline, Embase, Emcare, CINAHL, PsycInfo and Google Scholar in May to October 2022, screened 1663 articles and included 33 [2]. Two researchers assessed quality using the Medical Education Research Study Quality Instrument (MERSQI) [2]. A meta-analysis was not possible because study designs and outcomes varied [2].
- **AI in nursing.** This is an integrative review that used thematic synthesis and is indexed as a systematic review [3]. It included 25 studies and was guided by PRISMA 2020 and the SPIDER framework [3]. Study quality was assessed with the Mixed Methods Appraisal Tool (MMAT), and risk of bias with ROBINS-I [3]. The authors judged 21 of the 25 studies to be at moderate risk of bias [3].
- **Quality and safety in nursing.** A triservice military nursing team searched five databases and included 249 articles [4]. From these the team derived 94 practice recommendations and eight focus areas [4]. The work was intended to inform the quality and safety component of a nursing professional practice model for military hospitals [4].
- **WHO Surgical Safety Checklist.** This is a systematic review combining qualitative thematic analysis with meta-meta-analysis, meaning analysis of findings already reported in existing systematic reviews [5]. It drew on 20 systematic reviews [5]. Those reviews contained 24 unique observational cohort studies reporting pre-post data, that is, outcomes before and after checklist introduction, on 18 clinical outcomes [5].
- **Aficamten.** This was a phase 3, double-blind, randomized controlled trial in 282 adults with symptomatic obstructive hypertrophic cardiomyopathy (HCM), funded by Cytokinetics [1]. The authors describe elevated intracardiac pressure from left ventricular outflow tract obstruction as one of the major determinants of exercise intolerance in obstructive HCM [1]. Patients were assigned to aficamten (142) or placebo (140) for 24 weeks [1].

## Where the sources agree: workload as a shared concern

Workload appears in all four workforce and safety sources, but each treats it in a different role [2, 3, 4, 5]. In the burnout review, workload reduction is one of the organisational interventions, alongside job crafting and peer networks [2].

In the AI review, workload is a target for automation [3]. AI systems that automated routine follow-up tasks and generated predictive workload models were reported to free nurses from repetitive, non-clinical duties and to inform staffing decisions [3]. These efficiencies were associated with reductions in burnout and improved workplace morale [3]. The review reports this as an association drawn from studies mostly judged at moderate risk of bias, not as a demonstrated effect [3].

In the nursing quality and safety review, staffing, workload and work environment form one of the eight focus areas [4]. The other seven are communication, adverse events, leadership, patient experience, quality improvement, safety culture and committees, and technology and the electronic health record [4].

In the checklist meta-meta-analysis, workload appears as an aspect of using the tool rather than as its target [5]. One of six themes concerned the efficiency and workload involved in using the checklist [5]. The authors note that this theme is associated with checklist use but was not a focal area when the checklist was developed [5].

Taken together, the sources treat workload in three ways: as a target for intervention [2, 3], as a domain for practice recommendations [4], and as a theme in the experience of using a safety tool [5]. None of the abstracts reports a common workload measure, so the agreement is thematic rather than quantitative [2, 3, 4, 5].

## Where the sources pull in different directions

**Breadth of benefit.** The burnout review reports effective outcomes in 29 of 33 studies [2]. Those outcomes spanned improvements in well-being, work engagement, quality of life and resilience, and reductions in burnout, perceived stress, anxiety and depression [2]. The checklist meta-meta-analysis reaches a narrower conclusion [5]. It finds that the WHO SSC positively affects what it was explicitly designed to address and does not positively affect what it was not designed for [5]. The two findings concern different interventions and populations, so they do not contradict each other directly; one source reports benefit across many outcomes, while the other reports benefit confined to designed aims [2, 5].

**Clinical outcomes within the checklist evidence.** In the pooled pre-post studies, nine outcomes favored use of the WHO SSC: mortality, morbidity, surgical site infection, pneumonia, unplanned return to the operating room, urinary tract infection, blood loss requiring transfusion, unplanned intubation and sepsis [5]. Deep vein thrombosis was the only postoperative outcome assessed that did not favor the checklist [5]. Because the underlying studies were observational, these results indicate that checklist use was associated with better outcomes; they do not establish that it caused them [5].

**Benefits and costs of AI within one review.** The AI review itself contains opposing findings [3]. AI-powered simulations were linked to deeper student engagement, better case-management performance and higher satisfaction scores [3]. Learners also reported greater cognitive load and heightened stress during these activities [3]. AI alert algorithms and wearable sensors were reported to detect signs of patient deterioration and fever earlier than conventional methods [3]. Critical-care staff, however, stressed that automated insights must be balanced with professional judgment to avoid overreliance [3]. Nurses also raised ethical concerns about patient data privacy, algorithmic bias and the preservation of human-centered care [3].

**Where the evidence sits versus where recommendations point.** In the burnout review, 30 of 33 studies used individually focused interventions and only three were organisationally focused [2]. Similarly, 31 studies used secondary-level interventions, which manage stress in individuals, and only two used primary-level interventions, which eliminate the causes of stress [2]. Mindfulness-based practices were used in 20 studies [2]. The nursing quality and safety review, by contrast, places organisational domains such as leadership, safety culture and staffing among its eight focus areas [4]. The burnout evidence base is therefore concentrated on individual coping, while practice-model recommendations extend to the organisational level, where the burnout review found few studies [2, 4].

## How strong the evidence is

The aficamten trial reports the elements of a prespecified randomized design [1]. Its primary end point was change in peak oxygen uptake from baseline to week 24, measured by cardiopulmonary exercise testing [1]. Its 10 secondary end points were prespecified and tested hierarchically [1]. Peak oxygen uptake rose by a mean of 1.8 ml per kilogram per minute (95% CI, 1.2 to 2.3) with aficamten and by 0.0 ml per kilogram per minute (95% CI, -0.5 to 0.5) with placebo [1]. The least-squares mean between-group difference was 1.7 ml per kilogram per minute (95% CI, 1.0 to 2.4; P<0.001) [1]. All 10 secondary end points were significantly improved with aficamten compared with placebo [1]. The incidence of adverse events appeared similar between groups [1]. The trial population had a mean age of 59.1 years, 59.2% were men, and the baseline mean resting outflow tract gradient was 55.1 mm Hg [1].

The four workforce and safety abstracts report their evidence in less specific terms, and each in a different way [2, 3, 4, 5]. The checklist analysis is a meta-meta-analysis, but its clinical outcome data come from observational pre-post cohort studies, so its clinical findings rest on observational comparisons [5].

The burnout review states that the outcomes of numerous included studies were affected by design limitations, specifically no control or a waitlist control, and no post-intervention follow-up [2]. The abstract does not say how many of the 29 studies reporting effective outcomes were among those with these limitations [2].

The AI review judged most of its studies at moderate risk of bias [3]. Its authors still conclude that the evidence suggests improvements in critical thinking, learner engagement and clinical satisfaction [3]. Its finding on burnout is reported as an association [3].

The nursing quality and safety abstract reports the number of included articles and recommendations [4]. It does not report the designs of the included articles or how the evidence behind each recommendation was graded [4]. Readers therefore cannot judge the strength of individual recommendations from the abstract alone [4].

## What remains unknown

The sources leave several questions open [2, 3, 4, 5].

- Whether organisational, primary-level interventions perform better or worse than individual ones cannot be judged from the burnout review, which found only three organisationally focused studies [2].
- How long reported well-being gains last is unclear, because numerous burnout studies lacked post-intervention follow-up [2].
- The cost-effectiveness of AI in nursing is not established in the review abstract; its proposed Nursing AI Integration Roadmap (NAIIR) instead emphasizes rigorous economic evaluation [3].
- Whether the checklist associations would hold in other designs is not addressed, since the pooled clinical data are pre-post observational [5].
- How well recommendations developed for military hospitals transfer elsewhere is untested; the authors suggest the practices could be adapted to many healthcare delivery systems [4].
- Whether reducing workload through AI or organisational change improves patient safety outcomes, as distinct from staff well-being, is not reported in any of the abstracts [2, 3, 4].

## Limitations

This review is based on published abstracts, not full texts, so details of effect sizes, subgroup results and risk-of-bias judgments may be incomplete [1, 2, 3, 4, 5]. The five sources were not identified through a systematic search and cover different populations, settings and interventions, so their agreement is thematic and was not tested statistically [1, 2, 3, 4, 5]. The aficamten trial concerns a cardiac drug in patients with obstructive HCM and is included only as a methodological comparison, not as evidence about workforce interventions [1]. That trial was funded by Cytokinetics [1], and the nursing quality and safety review and the checklist analysis are both indexed as having non-U.S. government research support [4, 5]. The burnout review restricted inclusion to English-language full texts [2].

## References

1. Maron MS, Masri A, Nassif ME, et al. Aficamten for Symptomatic Obstructive Hypertrophic Cardiomyopathy. N Engl J Med. 2024;390(20):1849-1861. doi:10.1056/NEJMoa2401424. PMID: 38739079. https://pubmed.ncbi.nlm.nih.gov/38739079/
2. Cohen C, Pignata S, Bezak E, Tie M, Childs J. Workplace interventions to improve well-being and reduce burnout for nurses, physicians and allied healthcare professionals: a systematic review. BMJ Open. 2023;13(6):e071203. doi:10.1136/bmjopen-2022-071203. PMID: 37385740. https://pubmed.ncbi.nlm.nih.gov/37385740/
3. El Arab RA, Al Moosa OA, Sagbakken M, et al. Integrative review of artificial intelligence applications in nursing: education, clinical practice, workload management, and professional perceptions. Front Public Health. 2025;13:1619378. doi:10.3389/fpubh.2025.1619378. PMID: 40823249. https://pubmed.ncbi.nlm.nih.gov/40823249/
4. Patrician PA, Campbell CM, Javed M, et al. Quality and Safety in Nursing: Recommendations From a Systematic Review. J Healthc Qual. 2024;46(4):203-219. doi:10.1097/JHQ.0000000000000430. PMID: 38717788. https://pubmed.ncbi.nlm.nih.gov/38717788/
5. Sotto KT, Burian BK, Brindle ME. Impact of the WHO Surgical Safety Checklist Relative to Its Design and Intended Use: A Systematic Review and Meta-Meta-Analysis. J Am Coll Surg. 2021;233(6):794-809.e8. doi:10.1016/j.jamcollsurg.2021.08.692. PMID: 34592406. https://pubmed.ncbi.nlm.nih.gov/34592406/