Performance of artificial intelligence for the detection of pathological myopia from colour fundus images: a systematic review and meta-analysis

BACKGROUND

Pathological myopia (PM) is a major cause of worldwide blindness and represents a serious threat to eye health globally. Artificial intelligence (AI)-based methods are gaining traction in ophthalmology as highly sensitive and specific tools for screening and diagnosis of many eye diseases. However, there is currently a lack of high-quality evidence for their use in the diagnosis of PM.

METHODS

A systematic review and meta-analysis of studies evaluating the diagnostic performance of AI-based tools in PM was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidance. Five electronic databases were searched, results were assessed against the inclusion criteria and a quality assessment was conducted for included studies. Model sensitivity and specificity were pooled using the DerSimonian and Laird (random-effects) model. Subgroup analysis and meta-regression were performed.

RESULTS

Of 1021 citations identified, 17 studies were included in the systematic review and 11 studies, evaluating 165,787 eyes, were included in the meta-analysis. The area under the summary receiver operator curve (SROC) was 0.9905. The pooled sensitivity was 95.9% [95.5%-96.2%], and the overall pooled specificity was 96.5% [96.3%-96.6%]. The pooled diagnostic odds ratio (DOR) for detection of PM was 841.26 [418.37-1691.61].

CONCLUSIONS

This systematic review and meta-analysis provides robust early evidence that AI-based, particularly deep-learning based, diagnostic tools are a highly specific and sensitive modality for the detection of PM. There is potential for such tools to be incorporated into ophthalmic public health screening programmes, particularly in resource-poor areas with a substantial prevalence of high myopia.

Long Covid: conceptualizing the challenges for public health

BACKGROUND

Long Covid has caused significant disruption to public services, economies and population health worldwide, but no single public health approach has proven effective in its management. This essay was the winning entry for the Faculty of Public Health's Sir John Brotherston Prize 2022.

METHODS

In this essay, I synthesize existing literature on public health policy in long Covid, and discuss the challenges and opportunities posed by long Covid for the public health profession. The utility of specialist clinics and community care, in the UK and internationally, is examined, as well as key outstanding issues relating to evidence generation, health inequality and defining long Covid. I then use this information to inform a simple conceptual model.

RESULTS

The generated conceptual model integrates community- and population-level interventions; key areas of identified policy need at both levels include ensuring equitable access to long Covid care, developing screening programmes for high-risk populations, co-production of research and clinical services with patients, and using interventions to generate evidence.

CONCLUSIONS

Significant challenges remain in the management of long Covid from a public health policy perspective. Multidisciplinary community-level and population-level interventions should be employed with a view to achieving an equitable and scalable model of care.

Evaluation of prognostic risk models for postoperative pulmonary complications in adult patients undergoing major abdominal surgery: a systematic review and international external validation cohort study

BACKGROUND

Stratifying risk of postoperative pulmonary complications after major abdominal surgery allows clinicians to modify risk through targeted interventions and enhanced monitoring. In this study, we aimed to identify and validate prognostic models against a new consensus definition of postoperative pulmonary complications.

METHODS

We did a systematic review and international external validation cohort study. The systematic review was done in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. We searched MEDLINE and Embase on March 1, 2020, for articles published in English that reported on risk prediction models for postoperative pulmonary complications following abdominal surgery. External validation of existing models was done within a prospective international cohort study of adult patients (≥18 years) undergoing major abdominal surgery. Data were collected between Jan 1, 2019, and April 30, 2019, in the UK, Ireland, and Australia. Discriminative ability and prognostic accuracy summary statistics were compared between models for the 30-day postoperative pulmonary complication rate as defined by the Standardised Endpoints in Perioperative Medicine Core Outcome Measures in Perioperative and Anaesthetic Care (StEP-COMPAC). Model performance was compared using the area under the receiver operating characteristic curve (AUROCC).

FINDINGS

In total, we identified 2903 records from our literature search; of which, 2514 (86·6%) unique records were screened, 121 (4·8%) of 2514 full texts were assessed for eligibility, and 29 unique prognostic models were identified. Nine (31·0%) of 29 models had score development reported only, 19 (65·5%) had undergone internal validation, and only four (13·8%) had been externally validated. Data to validate six eligible models were collected in the international external validation cohort study. Data from 11 591 patients were available, with an overall postoperative pulmonary complication rate of 7·8% (n=903). None of the six models showed good discrimination (defined as AUROCC ≥0·70) for identifying postoperative pulmonary complications, with the Assess Respiratory Risk in Surgical Patients in Catalonia score showing the best discrimination (AUROCC 0·700 [95% CI 0·683-0·717]).

INTERPRETATION

In the pre-COVID-19 pandemic data, variability in the risk of pulmonary complications (StEP-COMPAC definition) following major abdominal surgery was poorly described by existing prognostication tools. To improve surgical safety during the COVID-19 pandemic recovery and beyond, novel risk stratification tools are required.

FUNDING

British Journal of Surgery Society.

Twelve tips for medical students to enhance clinical skills learning during disrupted placements

Disruption to healthcare settings can present a significant challenge to traditional, face-to-face models of learning in medical education. Reductions in undergraduate medical students' clinical exposure, whether due to periods of increased healthcare demand, localised service changes or infectious disease outbreaks, are likely to result in fewer opportunities to develop key clinical and practical skills. Proficiency in these skills is often essential to progression and future practice, creating a broad incentive for students to develop techniques to maintain and refine their clinical skills during disrupted placements. These tips, based on our experiences as senior medical students, are intended to help students on disrupted placements to engage in reflective practice, discover ways to facilitate further opportunities for clinical skills learning, and to make the most of clinical skills learning opportunities that they do receive.

Frameworks for implementation, uptake and use of digital health interventions in ethnic minority populations: a scoping review using cardiometabolic disease as a case study. (Preprint)

Background

Digital health interventions have become increasingly common across health care, both before and during the COVID-19 pandemic. Health inequalities, particularly with respect to ethnicity, may not be considered in frameworks that address the implementation of digital health interventions. We considered frameworks to include any models, theories, or taxonomies that describe or predict implementation, uptake, and use of digital health interventions.

Objective

We aimed to assess how health inequalities are addressed in frameworks relevant to the implementation, uptake, and use of digital health interventions; health and ethnic inequalities; and interventions for cardiometabolic disease.

Methods

SCOPUS, PubMed, EMBASE, Google Scholar, and gray literature were searched to identify papers on frameworks relevant to the implementation, uptake, and use of digital health interventions; ethnically or culturally diverse populations and health inequalities; and interventions for cardiometabolic disease. We assessed the extent to which frameworks address health inequalities, specifically ethnic inequalities; explored how they were addressed; and developed recommendations for good practice.

Results

Of 58 relevant papers, 22 (38%) included frameworks that referred to health inequalities. Inequalities were conceptualized as society-level, system-level, intervention-level, and individual. Only 5 frameworks considered all levels. Three frameworks considered how digital health interventions might interact with or exacerbate existing health inequalities, and 3 considered the process of health technology implementation, uptake, and use and suggested opportunities to improve equity in digital health. When ethnicity was considered, it was often within the broader concepts of social determinants of health. Only 3 frameworks explicitly addressed ethnicity: one focused on culturally tailoring digital health interventions, and 2 were applied to management of cardiometabolic disease.

Conclusions

Existing frameworks evaluate implementation, uptake, and use of digital health interventions, but to consider factors related to ethnicity, it is necessary to look across frameworks. We have developed a visual guide of the key constructs across the 4 potential levels of action for digital health inequalities, which can be used to support future research and inform digital health policies.

Supportive care needs of men with prostate cancer: A systematic review update

BACKGROUND

Prostate cancer is highly prevalent and impacts profoundly on patients' quality of life, leading to a range of supportive care needs.

METHODS

An updated systematic review and thematic synthesis of qualitative data using the Preferred Reporting Items for Systematic Reviews (PRISMA) reporting guidelines, to explore prostate cancer patients' experience of, and need for, supportive care. Five databases (Medline, Embase, PsycInfo, Emcare and ASSIA) were searched; extracted data were synthesised using Corbin and Strauss's 'Three Lines of Work' framework.

RESULTS

Searches identified 2091 citations, of which 105 were included. Overarching themes emerged under the headings of illness, everyday life and biographical work. Illness work needs include consistency and continuity of information, tailored to ethnicity, age and sexual orientation. Biographical work focused on a desire to preserve identity in the context of damaging sexual side effects. Everyday life needs centred around exercise and diet support and supportive relationships with partners and peers. Work-related issues were highlighted specifically by younger patients, whereas gay and bisexual men emphasised a lack of specialised support.

CONCLUSION

While demonstrating some overarching needs common to most patients with prostate cancer, this review offers novel insight into the unique experiences and needs of men of different demographic backgrounds, which will enable clinicians to deliver individually tailored supportive care.

Post-COVID-19 assessment in a specialist clinical service: a 12-month, single-centre, prospective study in 1325 individuals

Introduction

Post-COVID-19 complications require simultaneous characterisation and management to plan policy and health system responses. We describe the 12-month experience of the first UK dedicated post-COVID-19 clinical service to include hospitalised and non-hospitalised patients.

Methods

In a single-centre, observational analysis, we report the demographics, symptoms, comorbidities, investigations, treatments, functional recovery, specialist referral and rehabilitation of 1325 individuals assessed at the University College London Hospitals post-COVID-19 service between April 2020 and April 2021, comparing by referral route: posthospitalised (PH), non-hospitalised (NH) and post emergency department (PED). Symptoms associated with poor recovery or inability to return to work full time were assessed using multivariable logistic regression.

Results

1325 individuals were assessed (PH: 547, 41.3%; PED: 212, 16%; NH: 566, 42.7%). Compared with the PH and PED groups, the NH group were younger (median 44.6 (35.6–52.8) years vs 58.3 (47.0–67.7) years and 48.5 (39.4–55.7) years), more likely to be female (68.2%, 43.0% and 59.9%), less likely to be of ethnic minority (30.9%, 52.7% and 41.0%) or seen later after symptom onset (median (IQR): 194 (118–298) days, 69 (51–111) days and 76 (55–128) days; all p<0.0001). All groups had similar rates of onward specialist referral (NH 18.7%, PH 16.1% and PED 18.9%, p=0.452) and were more likely to require support for breathlessness (23.7%, 5.5% and 15.1%, p<0.001) and fatigue (17.8%, 4.8% and 8.0%, p<0.001). Hospitalised patients had higher rates of pulmonary emboli, persistent lung interstitial abnormalities and other organ impairment. 716 (54.0%) individuals reported <75% optimal health (median 70%, IQR 55%–85%). Less than half of employed individuals could return to work full time at first assessment.

Conclusion

Post-COVID-19 symptoms were significant in PH and NH patients, with significant ongoing healthcare needs and utilisation. Trials of interventions and patient-centred pathways for diagnostic and treatment approaches are urgently required.