According to World Health Organization estimates, approximately 19 million children under the age of 15 years are visually impaired, with 1.4 million exhibiting irreversible impairment.1 Among cases of severe visual impairment diagnosed before the age of 1 year, around one-third are attributable to genetic causes.2

Substantial proportions of childhood and adult-onset visual impairments are caused by inherited eye disorders (IEDs), which include anterior segment dysgenesis; inherited retinal disorders (IRDs); microphthalmia, anophthalmia, and coloboma; ocular tumours; congenital cataracts; and albinism. Over the past three decades, more than 450 genes have been associated with IEDs.2 3 Genetic diagnosis in such cases is challenging due to both clinical and genetic heterogeneity.

Ocular genetics has rapidly evolved over the past decade—from identifying inheritance patterns of IEDs to establishing genotype-phenotype correlations for disease prognostication and enrolling patients in gene therapy trials. In 2018, the United States Food and Drug Administration approved the first ocular gene therapy, Luxturna, for the treatment of RPE65-related inherited retinal disease.4 In 2012, the American Academy of Ophthalmology published diagnostic guidelines encouraging the routine use of genetic testing for IEDs.5 Multiple genes can now be assessed simultaneously through a single genomic test, which is particularly useful for identifying heterogeneous single-gene disorders and resolving cases where a clinical diagnosis is difficult to establish.6 Advances in sequencing technologies are uncovering the molecular aetiologies of various disorders. Consequently, the genomic approach to IEDs is gaining popularity, highlighting the need for more sophisticated genomic testing and comprehensive ocular genetic services.

In Hong Kong, the Retinitis Pigmentosa Registry—the first of its kind among Chinese populations globally—was established in 1995. Its main objectives are to provide detailed ophthalmic and genetic examinations for patients with inherited retinal degenerative diseases and to build a database for future scientific, medical, and sociological research.7 However, local data remain limited and the genetic landscapes of other IEDs are still unclear.

Hong Kong Children’s Hospital (HKCH) serves as the tertiary referral centre for complex, serious, and uncommon paediatric cases requiring multidisciplinary management, providing diagnosis, treatment, and rehabilitation services across the territory. In 2021, the Clinical Genetics Service Unit (CGSU) at HKCH was established as the first clinical genetics branch under the Hospital Authority. In July 2023, the Clinical Genetic Service (CGS) of the Department of Health (DH)—the former government-funded tertiary genetic referral centre providing genetic counselling and laboratory services to the entire Hong Kong population—was integrated with the CGSU and renamed the Department of Clinical Genetics (DCG) under the Hospital Authority. As a major clinical genetics service provider in Hong Kong, the DCG now offers genetic counselling services territory-wide.

Acknowledging the knowledge gap in the local genetic landscape and the lack of a comprehensive service model for patients with IEDs in Hong Kong, we conducted this retrospective review to analyse the local mutational spectrum across various IED subtypes and the corresponding diagnostic yield in our institution. Our aim was to better understand the clinical utility of genomic testing in IED patients and to formulate a comprehensive ocular genetic service model that addresses the needs of local patients.

Patients presenting with eye manifestations were retrospectively identified by querying records between 1 December 2021 and 30 October 2023 through the Hospital Authority Teams database under the CGSU/DCG at HKCH. The database included all patients who had attended genetic counselling clinics under the CGSU/DCG. Clinical geneticists and ophthalmologists reviewed all clinical notes, genetic reports, and electronic health records in the Clinical Management System, as well as paper records.

Patients’ phenotypes were reviewed and categorised by ophthalmologists into the following nine groups: (a) anterior segment dysgenesis; (b) IRDs; (c) cataract and lens disorders; (d) microphthalmia, anophthalmia, and coloboma spectrum; (e) neuro-ophthalmology (eg, optic atrophy); (f) ocular albinism or oculocutaneous albinism; (g) high myopia; (h) ocular tumours; and (i) others.

Patients with inconclusive eye phenotypes were excluded. Relevant history (including consanguinity, ethnicity, and family history), physical examination findings (dysmorphism and involvement of other systems), ophthalmological assessments and examinations, other relevant investigations (eg, magnetic resonance imaging of the brain and renal imaging), and previous genetic test reports were reviewed. A positive family history was defined as the presence of related eye phenotypes in a first-degree relative, or in two or more second- or third-degree relatives with the same condition.

All patients underwent comprehensive dysmorphology evaluations and genetic counselling, including pre-test and post-test consultations with the clinical genetics team. Prior to providing informed consent for genomic testing, patients were counselled on the indications, limitations, diagnostic yield, variants of uncertain clinical significance, and the ethical, social, and legal implications of genomic testing. Informed consent was obtained from affected patients or their legal guardians before undergoing diagnostic genomic testing.

According to clinical indications, patients were offered various genomic tests, including single-gene sequencing, array comparative genomic hybridisation, multiplex ligation–dependent probe amplification, whole-exome sequencing–based panels, medical exome sequencing, and mitochondrial sequencing. DNA was extracted from peripheral blood ethylenediaminetetraacetic acid samples. For mitochondrial sequencing, mitochondrial DNA extracted from urine-derived cells was used. All tests were performed in one of two accredited laboratories: the Genetic Laboratory of DH (which became a combined service with the Hospital Authority after July 2023), or the Genetics and Genomics Laboratory at HKCH, in accordance with laboratory-specific protocols and guidelines. Inheritance and phasing were determined via targeted Sanger sequencing of parental samples.

Clinical characteristics were collected from electronic records and, when available, hospital case notes and CGS paper records. These characteristics included age at onset, age at first encounter, sex, ethnicity, consanguinity, laterality of ocular involvement, severity of visual impairment, family history of ocular conditions, syndromic features, and other associated system involvement. Genetic testing results were retrieved from the Clinical Management System, CGS database, and paper records. Additionally, reproductive planning (for either the index patient or their parents) and other subspecialty referrals after a substantiated molecular diagnosis—as documented in genetic counselling notes—were recorded for clinical utility analysis. All clinical data are presented as percentages or means ± standard deviations, unless otherwise specified.

Molecular and clinical data from all recruited individuals were analysed using SPSS (Windows version 26.0; IBM Corp, Armonk [NY], United States). Categorical variables (eg, syndromic vs non-syndromic presentation, presence of family history) were compared using Fisher’s exact test, while continuous variables were compared using the independent samples t test. P values of less than 0.05 were considered statistically significant. This article was written in compliance with the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) reporting guidelines.

Between December 2021 and October 2023, 3653 patients were registered at the HKCH genetic counselling clinic. Of these, 148 symptomatic patients from 147 families met the inclusion criteria for this study. Approximately 4% of patients presented to the genetics clinic with ophthalmological diseases. Overall, 130 (87.8%) patients consented to genomic testing (Fig).

Among the 130 patients, approximately 92% were Chinese, with a male-to-female ratio of 3:2. The mean age (±standard deviation) at onset was 12.5±16.2 years. Within this cohort, 53.1% of patients were classified under IRDs, 14.6% under neuro-ophthalmology, and 13.8% under cataract/lens disorders.

Fifteen patients (11.5% of those tested) presented with more than one ocular phenotype. The majority of patients (>80%) exhibited bilateral ocular involvement. Detailed demographics, family history, and disease categories of the 130 patients who underwent genetic testing are presented in Table 1.

The diagnostic yield of genomic testing was defined as the proportion of individuals with pathogenic or likely pathogenic molecular variants or structural variants contributing to the clinical phenotype. A whole-exome sequencing–based virtual panel was requested for 78 (60%) of the 130 patients, based on their presenting phenotypes (online supplementary Table 1). Using this panel-based approach, the diagnostic yield was 51.3%. Twenty-three patients (17.7%) underwent single-gene testing based on highly specific phenotypes without molecular heterogeneity, such as RB1, CHD7, NF1, and RS1 (online supplementary Table 2). This single-gene approach successfully diagnosed 14 patients (60.8%). Medical exome sequencing was offered to 22 patients with multiple congenital anomalies or suspected syndromes, achieving a diagnostic yield of 50% (11/22). Two patients were diagnosed through copy number variation analysis (online supplementary Table 2).

The overall diagnostic yield for this cohort was 51.5% (Table 2). As mentioned earlier, 15 patients exhibited overlapping phenotypes across disease categories, with inherited retinal disorders and cataracts being the most common co-existing phenotypes. The microphthalmia, anophthalmia, and coloboma spectrum demonstrated the highest diagnostic yield at 100%. All five patients in this category presented with bilateral eye involvement and were syndromic (eg, two with CHARGE syndrome) [online supplementary Table 2]. Among the 69 IRD patients who underwent testing, 40 had confirmed molecular diagnoses, resulting in diagnostic yield of 58% for the IRD group. The most commonly identified genes were USH2A, ABCA4, COL2A1, RP1L1, and RS1 (online supplementary Table 2). No significant differences in diagnostic yield were detected across disease categories (Table 2).

Patients presenting with IRDs and neuro-ophthalmological conditions generally exhibited a later age at onset and age at first encounter compared with other categories, although these differences were not statistically significant (Table 2).

In total, 25 novel variants were identified in 25 patients across 20 genes. Of these, four remained of uncertain clinical significance despite further phasing and segregation analysis (online supplementary Table 2). Five variants were found in trans with another likely pathogenic variant in the same gene, consistent with autosomal recessive inheritance.8 Following detailed phenotypic correlation and variant curation, 16 previously unreported novel variants were confirmed to contribute to molecular diagnoses within this cohort.

A significant difference in the proportion of positive genetic test results was observed between patients with and without a family history of ocular conditions (P=0.037). However, among patients with a family history, the interval between symptom onset and the first visit to the genetics clinic was significantly longer. Positive molecular diagnoses were also more likely to be achieved in syndromic patients (P=0.0014) [Table 3].

As shown in Table 4, individuals with bilateral eye involvement had a greater proportion of positive genetic test results (54.9%), although this difference was not statistically significant (P=0.07). Additionally, no significant difference in diagnostic yield was observed according to age at onset (P=0.29).

Genomic testing is increasingly recognised as an important tool for establishing new diagnoses or confirming ones, particularly in the context of rare conditions, which are often complex and costly to diagnose, leading to prolonged diagnostic odysseys. Molecular findings may offer additional clinical utility, including: (1) avoidance of unnecessary investigations or treatments; (2) improved prognostic certainty or redirection of clinical care; (3) enhanced surveillance or timely referral for extraocular manifestations; (4) provision of pre-symptomatic or cascade testing for potentially affected family members; and (5) support for reproductive planning.

In total, 14 patients received revised diagnoses after genomic testing, representing 21% of positive cases (Table 5). These new diagnoses were related to syndromic conditions, such as CTNNB1-related neurodevelopmental disorders, or involved extraocular features, such as pantothenate kinase–associated neurodegeneration (online supplementary Table 2).

Through medical record review, we determined that approximately 10% of test-positive patients were able to avoid unnecessary investigations and treatments. In two cases, metabolic workups for congenital cataract were discontinued after diagnostic confirmation. One patient with a pathogenic ABCA4 variant was advised to withhold vitamin A supplementation. In another case, a syndromic diagnosis of SOX2-related microphthalmia eliminated the need for repeated magnetic resonance imaging of the brain and prompted clinicians to monitor for other potential systemic associations, enabling timely intervention. Overall, 74.6% of patients experienced at least one clinical benefit as a result of genomic testing. More than 70% of test-positive patients benefited from improved prognostic certainty or a redirection of care. Approximately 30% of patients—or their carrier or affected parents—were offered options for reproductive planning through either prenatal confirmatory testing or preimplantation genetic testing. Table 5 summarises the clinical and diagnostic utilities observed in this study.

In this cohort, we reviewed 130 patients who attended the HKCH genetic counselling clinic over a 23-month period. This review offers a snapshot of the local genomic landscape of IEDs. The overall diagnostic yield of molecular testing was 51.5%, which is comparable to previously reported yields, ranging from 25% to 70% depending on phenotype and testing methodology.6 9 10 11 12 13 14 15 16 17 18 19 20

Among IRDs, a highly heterogeneous group, the diagnostic yield was 58%. This finding is consistent with a recent systematic review which reported a yield of 61.3% (95% confidence interval=57.8%-64.7%) across 51 studies of mixed IRD phenotypes.21 Several studies have demonstrated that well-curated gene panels are as effective as medical exome sequencing in detecting pathogenic variants in patients with IRDs.16 19 20 21 22

In our cohort of ocular tumours, 29.4% of patients received germline molecular diagnoses; most of these patients had unilateral retinoblastoma with no family history. Neither routine next-generation sequencing nor Sanger sequencing is typically capable of detecting low-level mosaicism. A previous study reported germline RB1 mutation detection rates ranging from 10% to 55% in unilateral retinoblastoma, which are substantially lower than those observed in bilateral cases.23 In the present study, the oculocutaneous albinism/ocular albinism group had the lowest diagnostic yield at 25%. This low yield may be attributed to the small sample size and the predominance of ocular albinism cases, for which previous research has shown a considerably lower molecular diagnostic yield than oculocutaneous albinism.24

Four recurrent variants were identified in this cohort (online supplementary Table 2):

1. NM_000350.3 (ABCA4): c.1804C>T, p.(Arg602Trp). This variant is present at a very low frequency in the Genome Aggregation Database25 (gnomAD v2.1.1: 11 in 250 870 alleles), with a predominance in East Asian populations (gnomAD v2.1.1: 5 in 18 364 alleles). The exact carrier risk in our locality requires further research.
2. NM_000330.4 (RS1): c.214G>A, p.(Glu72Lys). A missense variant located in exon 4 of the RS1 gene. This variant is well documented in Chinese populations, where it accounts for 9.2% of variants in individuals with X-linked retinoschisis.26
3. NM_178857.6 (RP1L1): c.133C>T, p.(Arg45Trp). This hotspot mutation, located in exon 2 of RP1L1, is associated with occult macular dystrophy. Although its allele frequency is not particularly enriched in the Chinese population, it has been mentioned in case reports.27 28
4. NM_206933.3 (USH2A): c.5572+1G>A. A splice-site variant in intron 27 of the USH2A gene, which has been documented in the literature.29 It has a relatively high allele frequency in East Asians (gnomAD v2.1.1: 3 in 249 996 alleles; East Asian subset: 3 in 18 382 alleles).30

Another variant, NM_153638.4 (PANK2): c.655G>A, p.(Gly219Ser), is a rare missense variant absent from the general population. It was detected in our local database and reported in 2023.31 Neurodegeneration with brain iron accumulation 1A (OMIM #234200) is caused by biallelic pathogenic variants in PANK2. This rare condition is characterised by early-onset retinal degeneration or pigmented retinopathy, followed by subtle neurological deficits such as tremor and extrapyramidal symptoms. Both ocular and neurological features follow a progressive course. Notably, two unrelated patients in our database carried the same PANK2 variant. A large, population-based study is warranted to determine whether this variant represents a founder mutation in our locality.

Genomic testing has advanced considerably over the past decade. As next-generation sequencing technologies (eg, whole-genome sequencing and multi-omics analysis) become more prevalent, diagnostic yields continue to improve.32 Given the availability of existing therapies, such as voretigene neparvovec for RPE65-related diseases, clinical trials are increasingly investigating gene-based therapies, including gene replacement through viral vectors, mutation suppression via small molecules, and splice modulation using antisense oligonucleotides.33 34 In addition to ending the diagnostic odyssey, a molecular diagnosis informs clinical management, facilitates access to other clinical services, initiates surveillance for extraocular manifestations, and supports family planning.6 12 35

Disease prognostication is a key aspect of clinical utility, most commonly reported in the IRD group. For example, COL2A1-related Stickler syndrome carries a high risk of retinal detachment, which may be mitigated through prophylactic cryotherapy or laser retinopexy.36 37 38 Among patients with unilateral retinoblastoma, those harbouring germline variants require closer surveillance of the contralateral eye and enhanced vigilance for the potential development of other cancers later in life.39 40

Among the 67 test-positive patients, 35 were diagnosed with autosomal dominant conditions (Table 5), six of which were inherited from an affected parent. Approximately one-third of positive findings were attributed to autosomal recessive conditions, with both parents identified as heterozygous carriers. Nine patients had X-linked conditions; in nearly all cases, the mothers were confirmed as heterozygous carriers, except for two who declined genetic testing. In this context, molecular diagnosis is clearly beneficial for cascade screening and reproductive planning. In practice, however, the extent to which patients report these benefits is often influenced by age and family circumstances within the study cohort. As a result, direct comparisons of reported utility across studies remain challenging.

In Hong Kong, our genetic counselling department serves as the major referral centre, receiving patients from both public and private sectors. Individuals with more severe phenotypes are more likely to be referred, resulting in potential ascertainment bias.

Genomic testing was recommended by clinical geneticists based on the clinical phenotype. However, due to resource limitations, not all patients underwent the full spectrum of available tests, which may have resulted in an underestimation of the diagnostic yield. Additionally, certain clinical subgroups (eg, microphthalmia, anophthalmia, and coloboma) had limited sample sizes, potentially affecting diagnostic yield outcomes. Despite these limitations, this pilot study provides a reliable estimate of the mutational spectrum and diagnostic yield among local IED patients.

To our knowledge, this is the first retrospective study of IED patients to examine both the local genetic landscape and the clinical utility of genomic testing. Our findings highlight the importance of integrating modern genomic technologies into the management of patients with IEDs. They also underscore the need for an enhanced service model through a multidisciplinary team approach, implemented via a combined ocular genetics clinic.

Ideally, clinical utility should be assessed through a randomised controlled trial, which maximises internal validity and control for confounding variables. However, the level of evidence required varies according to clinical indication and type of genetic test. The data presented in this retrospective observational study, collected over nearly 2 years, are considered representative of real-world clinical scenarios. Future research involving multicentre collaborations over a longer period (eg, 10 years) will provide a more comprehensive understanding.

Interestingly, patients with a family history experienced a longer interval between symptom onset and their first encounter at the clinical genetics clinic (Table 3). This finding emphasises the importance of raising public awareness about the role of genomic medicine in managing IEDs.

The multidisciplinary team clinic model—comprising ophthalmologists, genetic counsellors, geneticists, and genetic nurses—is a current global trend for integrating genomic testing into clinical care pathways. It has been proven effective, particularly when applied to IRDs as a model.41 42 Similar models have also been adopted in other specialties, such as neurogenetics and cardiogenetics clinics.

At HKCH, a combined ocular genetics clinic commenced service in May 2022. The team includes ophthalmologists, genetic counsellors, clinical geneticists, optometrists, and nurses. Patients are referred from both public and private sectors for a variety of indications, such as atypical eye phenotypes, suspected syndromic conditions, or complex counselling needs (eg, variants of uncertain clinical significance detected in previous genomic tests conducted locally or overseas). This one-stop combined clinic enables joint discussions among specialists to formulate comprehensive management plans and reduces the need for repeated hospital visits, saving patients valuable time.

Approximately 4% of patients attending our genetic clinic had ocular disorders. The overall diagnostic yield of genomic testing was 51.5%; predominance was the strongest among patients with syndromic presentations and positive family history.

This study demonstrates high clinical utility of genomic testing in over 70% of patients with confirmed molecular diagnoses. There is a global shift towards managing IED patients through a multidisciplinary team clinic service model. To meet the growing demand for genomic medicine in IEDs, future studies should incorporate prospective, population-wide sampling, long-term follow-up, and multicentre collaboration.

Concept or design: SSW Cheng, HM Luk.
Acquisition of data: SSW Cheng.
Analysis or interpretation of data: SSW Cheng, SSL Cheung.
Drafting of the manuscript: All authors.
Critical revision of the manuscript for important intellectual content: All authors.

All authors had full access to the data, contributed to the study, approved the final version for publication, and take responsibility for its accuracy and integrity.

As an editor of the journal, JCS Yam was not involved in the peer review process. Other authors have disclosed no conflicts of interest.

The authors thank the patients and their families for contributing the clinical data used in this study.

Part of the research data was presented at the Kowloon Central Cluster Convention, 25 October 2024, Hong Kong.

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

This research was performed in compliance with the Declaration of Helsinki. Ethics approval was granted by the Hospital Authority Central Institutional Review Board, Hong Kong (Ref No.: PAED-2023-076). A waiver of patient consent was obtained from the Committee due to the retrospective nature of the research.

The supplementary material was provided by the authors and some information may not have been peer reviewed. Accepted supplementary material will be published as submitted by the authors, without any editing or formatting. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by the Hong Kong Academy of Medicine and the Hong Kong Medical Association. The Hong Kong Academy of Medicine and the Hong Kong Medical Association disclaim all liability and responsibility arising from any reliance placed on the content.

1. National Academies of Sciences, Engineering, and Medicine; Health and Medicine Division; Board on Health Care Services; Board on the Health of Select Populations; Committee on the Evidence Base for Genetic Testing. An Evidence Framework for Genetic Testing. Washington (DC): National Academies Press (US); 2017.

2. Rahi JS, Cable N; British Childhood Visual Impairment Study Group. Severe visual impairment and blindness in children in the UK. Lancet 2003;362:1359-65. Crossref

3. Chen HY, Lehmann OJ, Swaroop A. Genetics and therapy for pediatric eye diseases. EBioMedicine 2021;67:103360. Crossref

4. FDA approves hereditary blindness gene therapy. Nat Biotechnol 2018;36:6. Crossref

5. Stone EM, Aldave AJ, Drack AV, et al. Recommendations for genetic testing of inherited eye diseases: report of the American Academy of Ophthalmology task force on genetic testing. Ophthalmology 2012;119:2408-10. Crossref

6. Burdon KP. The utility of genomic testing in the ophthalmology clinic: a review. Clin Exp Ophthalmol 2021;49:615-25. Crossref

7. Lam ST, To CH, Leung KW, Yip SP, Lo IF, Tsang KP. Lessons learnt from a genetic disease registry in Hong Kong. Hong Kong Med J 2021;27:226-8. Crossref

8. Richards S, Aziz N, Bale S, et al. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med 2015;17:405-24. Crossref

9. Ma A, Grigg JR, Flaherty M, et al. Genome sequencing in congenital cataracts improves diagnostic yield. Hum Mutat 2021;42:1173-83. Crossref

10. Haug P, Koller S, Maggi J, et al. Whole exome sequencing in coloboma/microphthalmia: identification of novel and recurrent variants in seven genes. Genes (Basel) 2021;12:65. Crossref

11. García Bohórquez B, Aller E, Rodríguez Muñoz A, Jaijo T, García García G, Millán JM. Updating the genetic landscape of inherited retinal dystrophies. Front Cell Dev Biol 2021;9:645600. Crossref

12. Lenassi E, Clayton-Smith J, Douzgou S, et al. Clinical utility of genetic testing in 201 preschool children with inherited eye disorders. Genet Med 2020;22:745-51. Crossref

13. Wang P, Li S, Sun W, et al. An ophthalmic targeted exome sequencing panel as a powerful tool to identify causative mutations in patients suspected of hereditary eye diseases. Transl Vis Sci Technol 2019;8:21. Crossref

14. Patel A, Hayward JD, Tailor V, et al. The oculome panel test: next-generation sequencing to diagnose a diverse range of genetic developmental eye disorders. Ophthalmology 2019;126:888-907. Crossref

15. Martin-Merida I, Avila-Fernandez A, Del Pozo-Valero M, et al. Genomic landscape of sporadic retinitis pigmentosa: findings from 877 Spanish cases. Ophthalmology 2019;126:1181-8. Crossref

16. Wang L, Zhang J, Chen N, et al. Application of whole exome and targeted panel sequencing in the clinical molecular diagnosis of 319 Chinese families with inherited retinal dystrophy and comparison study. Genes (Basel) 2018;9:360. Crossref

17. Lasseaux E, Plaisant C, Michaud V, et al. Molecular characterization of a series of 990 index patients with albinism. Pigment Cell Melanoma Res 2018;31:466-74. Crossref

18. Haer-Wigman L, van Zelst-Stams WA, Pfundt R, et al. Diagnostic exome sequencing in 266 Dutch patients with visual impairment. Eur J Hum Genet 2017;25:591-9. Crossref

19. Saudi Mendeliome Group. Comprehensive gene panels provide advantages over clinical exome sequencing for Mendelian diseases. Genome Biol 2015;16:134. Crossref

20. Consugar MB, Navarro-Gomez D, Place EM, et al. Panel-based genetic diagnostic testing for inherited eye diseases is highly accurate and reproducible, and more sensitive for variant detection, than exome sequencing. Genet Med 2015;17:253-61. Crossref

21. Britten-Jones AC, Gocuk SA, Goh KL, Huq A, Edwards TL, Ayton LN. The diagnostic yield of next generation sequencing in inherited retinal diseases: a systematic review and meta-analysis. Am J Ophthalmol 2023;249:57-73. Crossref

22. Hayman T, Millo T, Hendler K, et al. Whole exome sequencing of 491 individuals with inherited retinal diseases reveals a large spectrum of variants and identification of novel candidate genes. J Med Genet 2024;61:224-31. Crossref

23. Gupta H, Malaichamy S, Mallipatna A, et al. Retinoblastoma genetics screening and clinical management. BMC Med Genomics 2021;14:188. Crossref

24. Chan KS, Bohnsack BL, Ing A, et al. Diagnostic yield of genetic testing for ocular and oculocutaneous albinism in a diverse united states pediatric population. Genes (Basel) 2023;14:135. Crossref

25. Genome Aggregation Database. Available from: https://gnomad.broadinstitute.org/. Accessed 1 Sep 2024.

26. Huang L, Sun L, Wang Z, et al. Clinical manifestation and genetic analysis in Chinese early onset X-linked retinoschisis. Mol Genet Genomic Med 2020;8:e1421. Crossref

27. Qi YH, Gao FJ, Hu FY, et al. Next-generation sequencing–aided rapid molecular diagnosis of occult macular dystrophy in a Chinese family. Front Genet 2017;8:107. Crossref

28. Xiao S, Sun W, Xiao X, et al. Clinical and genetic features of retinoschisis in 120 families with RS1 mutations. Br J Ophthalmol 2023;107:367-72. Crossref

29. Lin YW, Huang YS, Lin CY, et al. High prevalence of exon-13 variants in USH2A-related retinal dystrophies in Taiwanese population. Orphanet J Rare Dis 2024;19:238. Crossref

30. Chau JF, Yu MH, Chui MM, et al. Comprehensive analysis of recessive carrier status using exome and genome sequencing data in 1543 Southern Chinese. NPJ Genom Med 2022;7:23. Crossref

31. Wong EW, Cheng SS, Woo TT, Lam RF, Lai FH. Concurrent PANK2 and OCA2 variants in a patient with retinal dystrophy, hypopigmented irides and neurodegeneration. Ophthalmic Genet 2023;44:403-7. Crossref

32. Weisschuh N, Mazzola P, Zuleger T, et al. Diagnostic genome sequencing improves diagnostic yield: a prospective single-centre study in 1000 patients with inherited eye diseases. J Med Genet 2024;61:186-95. Crossref

33. Drag S, Dotiwala F, Upadhyay AK. Gene therapy for retinal degenerative diseases: progress, challenges, and future directions. Invest Ophthalmol Vis Sci 2023;64:39. Crossref

34. Tan F, Li X, Wang Z, Li J, Shahzad K, Zheng J. Clinical applications of stem cell–derived exosomes. Signal Transduct Target Ther 2024;9:17. Crossref

35. Sahu A, Kaur S, Sukhija J, Srivastava P, Kaur A. Spectrum of congenital and inherited ocular disorders seen in a genetic clinic: experience of a developing ocular genetic service. Indian J Ophthalmol 2023;71:935-40. Crossref

36. Fincham GS, Pasea L, Carroll C, et al. Prevention of retinal detachment in Stickler syndrome: the Cambridge prophylactic cryotherapy protocol. Ophthalmology 2014;121:1588-97. Crossref

37. Savarirayan R, Bompadre V, Bober MB, et al. Best practice guidelines regarding diagnosis and management of patients with type II collagen disorders. Genet Med 2019;21:2070-80. Crossref

38. Khanna S, Rodriguez SH, Blair MA, Wroblewski K, Shapiro MJ, Blair MP. Laser prophylaxis in patients with Stickler syndrome. Ophthalmol Retina 2022;6:263-7. Crossref

39. Tonorezos ES, Friedman DN, Barnea D, et al. Recommendations for long-term follow-up of adults with heritable retinoblastoma. Ophthalmology 2020;127:1549-57.Crossref

40. Abramson DH. Re: Skalet et al.: Screening children at risk for retinoblastoma: consensus report from the American Association of Ophthalmic Oncologists and Pathologists (Ophthalmology. 2018;125:453-458). Ophthalmology 2018;125:e63-4. Crossref

41. Davison N, Payne K, Eden M, et al. Exploring the feasibility of delivering standardized genomic care using ophthalmology as an example. Genet Med 2017;19:1032-9. Crossref

42. Black GC, Sergouniotis P, Sodi A, et al. The need for widely available genomic testing in rare eye diseases: an ERN-EYE position statement. Orphanet J Rare Dis 2021;16:142. Crossref

The Hong Kong College of Orthopaedic Surgeons (HKCOS) is the official professional body responsible for organising the training of orthopaedic surgeons and conducting specialist examinations in orthopaedics in Hong Kong. The College also plays a key role in setting standards of care in orthopaedic surgery, which includes, but is not limited to, the investigation, preservation, and restoration of the form and function of the extremities, spine, and associated structures. The aim is to bring relief to patients of all ages affected by injury or disease of the musculoskeletal system.1

Knee osteoarthritis (knee OA) is common among middle-aged and older individuals. Its incidence and severity increase with age.2 Hong Kong, one of the cities with the highest life expectancy globally,3 faces a growing demand for knee OA treatment.4 In the public healthcare sector, patients often experience a long waiting time for medical consultation, physiotherapy, and surgical interventions related to knee OA.4 With an ageing population, the number of individuals affected by knee OA is expected to further increase over the next two to three decades,5 highlighting the urgent need for effective management strategies in Hong Kong.

To provide evidence-based guidance for the public and healthcare professionals on the management of knee OA, the HKCOS published a position statement in August 2022 (Table 1).6 This position statement was developed based on a consensus survey conducted in June 2022, along with a review of high-quality evidence in the literature. A total of 26 treatment recommendations were formulated, reflecting the consensus of expert orthopaedic surgeons in Hong Kong. The methodology used to develop this position statement is available on the College’s website.6

To gather updated expert opinions on the management of knee OA, the College conducted a second survey in January 2024. This article presents the findings from the 2022 and 2024 surveys.

This study was conducted under the supervision and ethical oversight of the Council of HKCOS. The requirement to obtain informed consent from participants was waived by the Council. Two cross-sectional surveys were conducted online in June 2022 and January 2024, respectively. The same online questionnaire was used for both study periods to ensure consistency in data collection.

The online questionnaire was developed by the HKCOS Osteoarthritis of the Knee Working Group (HKCOS-OAKWG).6 A total of 26 interventions for knee OA were selected based on the practices of orthopaedic surgeons in Hong Kong and guidelines from international recognised organisations.7 8 An independent literature review was conducted, and the quality of evidence was graded according to the Oxford Centre for Evidence-Based Medicine levels of evidence.9 Available Level 1 and Level 2 evidence up to March 2022 was summarised for each intervention.

Fellows of the HKCOS listed in the Specialist Register (Orthopaedics and Traumatology) of the Hong Kong Medical Council were invited via email to participate in the surveys. Each survey remained open for 4 weeks. The surveys collected opinions on the 26 interventions. For each item, the questionnaire provided: (1) a brief description of the intervention; (2) a concise summary of the supporting Level 1 and Level 2 evidence; and (3) a single question asking whether the expert would recommend the intervention. Expert panellists were instructed to select one of three response options: ‘Recommend’, ‘Do not recommend’, or ‘Abstain’.

Descriptive statistics were reported for the two cross-sectional surveys conducted in 2022 and 2024. These included the number of expert panellists who participated in each survey with their recommendations regarding the 26 interventions for knee OA. The number of panellists who participated in both the 2022 and 2024 surveys was also recorded. The experience of the expert panel was described based on the number of years each panellist had been registered as a specialist with the Hong Kong Medical Council. Consensus was defined as a minimum of 70% agreement among the expert panellists. The resulting recommendations formed the basis of the HKCOS position statement. No analytical statistical tests were performed.

A total of 106 expert panellists participated in the 2022 survey, while 28 participated in the 2024 survey. At the time of survey distribution in June 2022 and January 2024, the College had 528 and 567 fellows, respectively. In June 2022, the numbers of active fellows, overseas fellows, and inactive fellows were 495, 4, and 27, respectively. The corresponding numbers in January 2024 were 526, 7, and 34.10 The response rates for the surveys were 20% in 2022 and 5% in 2024, respectively. Twenty expert panellists participated in both the 2022 and 2024 surveys. The mean duration of experience as registered orthopaedic specialists with the Hong Kong Medical Council (±standard deviation) was 15.6±9 years for the 2022 cohort and 19.2±7 years for the 2024 cohort (Fig 1). The distribution of expert panellists according to their affiliations is presented in Figure 2.

The results of the expert recommendations from the 2022 and 2024 surveys are summarised in Table 2. Consensus was reached among the expert panellists for 14 interventions in 2022 and 13 interventions in 2024. As consensus had already been achieved for most treatment options in the first-round survey, no changes were made to the HKCOS position statement on the management of knee OA, and the content of the second survey remained unchanged. In both surveys, a supermajority of panellists (≥70%) recommended 11 interventions: patient education; land-based exercise; water-based exercise; self-management programme; weight reduction; use of a cane; oral paracetamol; topical non-steroidal anti-inflammatory drugs (NSAIDs); oral NSAIDs; high tibial osteotomy; and joint replacement surgery. Conversely, at least 70% of respondents in both surveys did not recommend the use of lateral wedge insoles or denervation therapy. Consensus was not achieved for 12 interventions in 2022 and 13 interventions in 2024.

A consensus of 70% or greater agreement was reached among the expert panellists for 14 interventions in the 2022 survey and 13 interventions in the 2024 survey. Consensus remained consistent across both cross-sectional surveys for 13 interventions, including: recommendations in favour of patient education, land-based and water-based exercise, self-management programmes, weight reduction, use of a cane, oral paracetamol, topical and oral NSAIDs, high tibial osteotomy and joint replacement surgery, as well as recommendations against the use of lateral wedge insoles and denervation therapy.

There are two primary objectives in the management of knee OA: (1) control of disease progression; and (2) control of symptoms. Regarding disease control, it is rare to reverse the underlying pathological changes, such as joint space narrowing and varus malalignment of the lower limb, through non-operative treatment once the condition is established. The aim of management is to slow the progression of osteoarthritis by promptly identifying and treating ‘joint-threatening’ pathologies (such as symptomatic loose bodies, full-thickness, full-width root tears of the meniscus, and spontaneous osteonecrosis of the knee)6 11 and by controlling risk factors associated with disease progression (eg, through weight reduction and prevention of knee injury).12 Regarding symptom control, most non-operative treatments focus on pain management13 and improving quality of life.14 Pain is the most common presenting symptom in knee OA.15 However, there is only a weak correlation between radiological abnormalities and patient-reported symptoms.16 The severity of pain experienced is influenced by sensitisation mechanisms—including both peripheral and central sensitisation—rather than by anatomical changes in the arthritic joint alone.15 It is increasingly recognised that inflammation plays a prominent role in the pathogenesis and symptoms of OA.17 Inflammatory cytokines (eg, interleukin 6 and tumour necrosis factor alpha) and inflammatory mediators (eg, prostaglandin and bradykinin) contribute to peripheral sensitisation of nociceptors in the subchondral plate and joint capsule by activating G protein–coupled receptors, ionotropic receptors, and tyrosine kinase receptors located on nerve terminals and cell bodies.18 Accordingly, most treatments for knee OA aim to control inflammation in order to alleviate symptoms.7 The cell bodies of nociceptors are located in the dorsal root ganglia of the spinal cord, where nociceptive input is transmitted to the brain and brainstem via the spinothalamic tract, spinoreticular tract, spinomesencephalic tract, and spinohypothalamic tract.13 Persistent nociceptive stimulation leads to central sensitisation, characterised by hyperactivity and hyperexcitability of neurons in the brain and spinal cord.19 Inflammation also contributes to central sensitisation by increasing the production of neurotransmitters and neuromodulators, including glutamate and substance P.18 Furthermore, there is a loss of inhibitory control due to a reduction in inhibitory neurotransmitters, primarily gamma-aminobutyric acid (GABA), resulting in heightened pain sensitivity.18 The perception of nociceptive stimuli as pain is further influenced by psychological factors (such as depression, anxiety, and poor coping skills) and social factors (such as workers’ compensation claims and lack of social support). The interaction among biological, psychological, and social factors is described by the biopsychosocial model, which is considered important in the development of chronic pain.19 As a result, education, self-management programmes, and cognitive behavioural therapy are also important treatment strategies for knee OA.7

Surgical treatments, such as joint replacement surgery and high tibial osteotomy, are typically reserved for patients with end-stage knee arthritis or those with persistent symptoms despite adequate non-operative management.14

Most patient education programmes provide guidance on: (1) the nature, presenting symptoms, and treatment options for knee OA; (2) the importance of adherence to treatment and the adoption of preventive lifestyle behaviours; and (3) the expected natural disease course and treatment outcomes.20 Patient education operates through a theory of planned behaviour-based intervention.20 By modifying patients’ attitudes and the subjective norms of family members, both symptoms and quality of life can be improved, even in the absence of changes in physical condition.20

There is high-quality evidence in the literature supporting the effectiveness of exercise—both land-based and water-based—along with weight reduction, use of a cane, and self-management programmes in reducing pain and improving quality of life in patients with knee OA.21

Exercise alleviates symptoms of knee OA by increasing lower limb muscle strength and improving overall physical fitness.21 Improved muscle strength alters joint biomechanics, thereby reducing joint loading and pain. Aerobic exercise increases peak oxygen uptake by muscles and enhances the patient’s overall fitness, facilitating the performance of daily activities and enhancing functional ability.21 Exercise should be performed regularly, with sufficient—but not excessive—intensity.14 Recommended land-based activities include lower limb strengthening exercises, static cycling, walking programmes, Tai Chi, and Baduanjin,21 while water-based activities include swimming and pool therapy.22

Obesity is a known risk factor for the development of knee OA12 and the subsequent need for joint replacement surgery.23 Beyond increased mechanical loading, obesity is associated with metabolic syndrome—a cluster of metabolic disorders including abdominal obesity, hyperlipidaemia, hypertension, and elevated fasting serum glucose.24 Metabolic syndrome is increasingly recognised as an important risk factor for OA due to its association with systemic inflammation.25 Among obese individuals with OA, disability may be reduced with weight loss exceeding 5%.26 Evidence supports the use of intensive low-energy diet programmes for achieving weight loss and reducing pain in this population.27

The use of a cane reduces weight-bearing on the arthritic joint, thereby alleviating pain. A randomised controlled trial has demonstrated significant improvements in pain and function among patients who use a cane for walking.28

Self-management programmes have been shown to reduce pain and improve function in patients with knee OA.29 30 31 In addition to education and exercise, these programmes equip patients with a range of coping and management strategies, including pain coping techniques,29 joint protection and injury prevention strategies,30 and stress management skills.31

Pain may arise when nociceptors in the knee joint are stimulated.13 Although articular cartilage and the meniscus are commonly damaged in knee OA, they are aneural and therefore not direct sources of pain. In contrast, nociceptors are present in the subchondral bone, periosteum, joint capsule, and infrapatellar fat pad.13 In a healthy joint, these nociceptors are not stimulated during physiological loading, meaning that individuals without OA do not experience pain during daily activities such as walking, squatting, or running. Inflammatory cytokines, including interleukin 6, tumour necrosis factor alpha, and other pro-inflammatory molecules (eg, prostaglandin and bradykinin), increase in response to joint injury or during OA flare-ups. These molecules bind to nociceptors on the cell surface, inducing hypersensitivity and hyperexcitability. Consequently, pain may be experienced even during normal joint loading in previously painless knees—a phenomenon known as peripheral sensitisation.18 Non-steroidal anti-inflammatory drugs are effective in alleviating pain in knee OA32 by reducing intra-articular and circulating levels of inflammatory molecules, including prostaglandins.

While oral NSAIDs are effective in relieving OA-related pain, their side-effects warrant caution.33 Systemic NSAIDs should be used judiciously in older patients and in individuals with peptic ulcer disease, renal impairment, or a history of asthma or ischaemic heart disease.34 Topical NSAIDs, by contrast, have demonstrated efficacy in reducing knee pain while offering a much safer side-effect profile.35 As such, topical NSAIDs should be considered before initiating oral formulations.34

Paracetamol has been shown to improve pain and function in patients with knee OA.36 It is considered a safe medication, even in patients at higher risk of complications.7 However, overdose can result in hepatotoxicity and death, necessitating proper patient education regarding safe dosing practices.8

Both joint replacement surgery and high tibial osteotomy are salvage operations for symptomatic end-stage knee arthritis following unsuccessful non-operative treatment. Both procedures have demonstrated effectiveness in relieving pain in patients with knee OA.37 38 Joint replacement surgery has been shown to improve function regardless of the extent of joint involvement.39 In contrast, high tibial osteotomy is effective for symptom relief only in a selected group of patients whose symptomatic OA is confined to the medial compartment of the knee.38

The recommendations of the HKCOS position statement on the management of knee OA,6 alongside those of the National Institute for Health and Care Excellence (NICE) guideline,14 the American Academy of Orthopaedic Surgeons (AAOS) guideline,7 and the Osteoarthritis Research Society International (OARSI) guideline,33 are summarised in Table 3.

All of these guidelines recommend patient education, exercise, self-management programmes, weight reduction, and the use of a cane as appropriate management strategies for patients with knee OA.6 7 14 33 With respect to knee biomechanical interventions, such as knee braces and sleeves, the AAOS guideline is unique in suggesting that these “may be used” in knee OA.7 Concerning foot biomechanical interventions, the HKCOS, NICE, and AAOS guidelines advise against the use of lateral wedge insoles.6 7 14

For pharmacological treatment, all reviewed guidelines recommend the use of NSAIDs, either in topical or oral formulations.6 7 14 33 However, owing to the increased risk of adverse effects associated with oral NSAIDs in patients with co-morbidities, both HKCOS and NICE recommend topical NSAIDs as first-line treatment, reserving oral NSAIDs for second-line use.6 14

There is variation in guidance concerning the use of oral paracetamol. While HKCOS and AAOS support its use as a first-line treatment in knee OA,6 7 NICE and OARSI do not recommend its routine use.14 33 None of these guidelines recommend opioids for the treatment of knee OA.6 7 14 33 Regarding intra-articular steroid injections, the NICE, AAOS, and OARSI guidelines either recommend or conditionally recommend their use for short-term symptom relief.7 14 33

In contrast, intra-articular hyaluronic acid injections are not recommended in the NICE and AAOS guidelines.7 14 The only reviewed guideline that conditionally supports their use is the OARSI guideline.33

Based on the comparison outlined above, substantial differences are observed among the various guidelines (Table 3), despite all having evaluated the same set of treatment options for knee OA and drawing upon similar English-language literature. These differences may reflect variations in cultural context and healthcare systems across different geographical regions. Hong Kong is distinct from North America, Europe, and the United Kingdom due to the considerable influence of traditional Chinese medicine (TCM) within its healthcare system. Many patients in Hong Kong seeking orthopaedic care for knee OA—whether through the Hospital Authority or private healthcare providers—also receive TCM treatment. Some doctors registered with the Medical Council of Hong Kong have received training in TCM, and basic education in TCM is included in the Bachelor of Medicine and Bachelor of Surgery curriculum in Hong Kong. It is thus crucial for the HKCOS to develop a position statement tailored to the local population. The comparison of various guidelines highlights differences between the HKCOS position statement and guidelines issued by other reputable international organisations. The recommendations put forward by HKCOS may be better suited to the needs of patients in Hong Kong. This position statement also serves a valuable reference for healthcare systems in which TCM plays a prominent role, including those in Mainland China, Taiwan, Korea, and Japan.

The present study has several limitations. First, its descriptive design precluded the use of analytical statistics; as a result, we were unable to identify whether there were significant shifts in opinion regarding interventions for knee OA within the Hong Kong orthopaedic community between 2022 and 2024. Second, only 19% of the expert panellists who participated in the 2022 survey also took part in the 2024 study. Consequently, we could not assess potential changes in opinion among those who did not respond in 2024. Third, the overall response rate for the 2024 survey was low, which may limit the representativeness of the findings and reduce their generalisability to the wider orthopaedic specialist population in Hong Kong. Lastly, the threshold for consensus, defined as a minimum of 70% agreement among the expert panellists, was arbitrary and may not be universally accepted.

Findings from both the 2022 and 2024 surveys demonstrated consistent expert consensus on key recommended and non-recommended interventions for knee OA. The recommendations provide evidence-based guidance to the public and medical practitioners regarding the effectiveness of various management strategies for knee OA.

The author solely contributed to the concept or design of the study, acquisition of the data, analysis or interpretation of the data, drafting of the manuscript, and critical revision of the manuscript for important intellectual content.

The author had full access to the data, contributed to the study, approved the final version for publication, and takes responsibility for its accuracy and integrity.

The author has disclosed no conflicts of interest.

The author would like to acknowledge the contributions of the following Fellows of the Hong Kong College of Orthopaedic Surgeons, who served as members of the expert panel and provided recommendations on interventions for the management of knee osteoarthritis in the survey conducted in 2022 (in alphabetical order):
Chan Baldwin, Chan Ka-wah, Chan Sai-keung, Chan Shiuwai, Chan Tsang-tung, Chan Tze-wang, Chan Wai-kwan Vincent, Chan Wai-lam, Chan Wai-sing, Chang His-tse Joseph Jeremy, Chang Shao, Chang Yun-po Robert, Cheng Hung-on, Cheong Peng-meng, Cheung Chi-nok, Cheung Ho-man, Cheung Man-hong, Cheung Pui-yin Jason, Cheung Wai-yuen, Cheung Wang-yan Warren, Cheung Yim-ling Amy, Chien Ping Eric, Chiu Chi-kit, Chiu Kwong-yuen Peter, Choi Shing-hing, Choi Sum-hung, Chow Kai-pun, Chu Kai-man, Chun Siu-yeung, Chung Kwong-yin, Chung Man-ting Marvin, Fan Chi-ho Jason, Fang Christian Xinshuo, Fu Chun-him Henry, Fu Wai-kee, Fung Kwai-yau, Hau Vincent, Ho Ki-wai, Ho Sheung-tung, Hung Leong-pan, Kwok Hau-yan, Lai Kam-keung, Lau Chi-yuen, Lau Chi-yuk, Lau Hoi-kuen, Lau Tak-wing, Lau Yip-kwong Francis, Law Sheung-wai, Law Ying-kan, Lee Qunn-jid, Lee Sung-yee, Lee Yeung-fai, Leung Ka-hei, Leung Yum-kwong, Ling Ka-kin Samuel, Loong Tak-wan, Lui Ming-yan, Man Lok-pong, Ng Fu-yuen, Ng Cheukkee Kenneth, Ng Pak-lin Eugene, Ng Wai-kit Raymond, Ng Weng-io, Ngai Wai-kee, Ong Tim-yun Michael, Sin Cheuk-hang, Siu Kam-to, Sun Lun-kit, Sun Kin-wai Kelvin, Tam Ka-ki, Tam Kwok-wai Kelvin, Tang Yan-ho Bruce, Tang Yuk-kwan, Tong Hoi-yiu Sara, Tse Lung-fung, Tsoi Chi-wah Danny, Tung Kam-lung, Wong Hok-leung, Wong Kwok-shing Patrick, Wong Man-kwan, Wong Nang-man Raymond, Wong Sze-hung, Wu Daniel Yiang, Wu Tsz-kit, Yang Isaac Bruce, Yeung Lok-yin Michael, Yeung Sze-tsun Eric, Yeung Yip-kan, Yuen Shiu-him Jonathan, Yung Shing-yat Colin.

The author would like to acknowledge the contributions of the following Fellows of the Hong Kong College of Orthopaedic Surgeons, who served as members of the expert panel that provided recommendations on interventions for the management of knee osteoarthritis in the 2024 survey (in alphabetical order):
Chan Shiu-wai, Chan Wai-lam, Cheung Chi-nok, Cheung Wang-yan Warren, Chiu Kwong-yuen Peter, Choi Shing-hing, Fu Chun-him Henry, Ho Sheung-tung, Lam Chi-keung Johnson, Lau Pui-yau, Lee Qunn-jid, Leung Ka-hei, Liem Man-shing, Ng Fu-yuen, Wong Hok-leung, Yeung Sai-hung.

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

This research was approved by the Council of the Hong Kong College of Orthopaedic Surgeons. The requirement for informed consent was waived by the Council.

1. The Hong Kong College of Orthopaedic Surgeons. Constitution. Available from: https://www.hkcos.org.hk/aboutus_constitution.asp. Accessed 4 Jul 2022.

2. Tang X, Wang SF, Zhan SY, et al. The prevalence of symptomatic knee osteoarthritis in China. Results from the China Health and Retirement Longitudinal Study. Arthritis Rheumatol 2016;68:648-53. Crossref

3. Woo J. Age and Ageing journal 50th anniversary commentary series. Healthcare for older people in Asia. Age Ageing 2022;51:afab189. Crossref

4. Hospital Authority, Hong Kong SAR Government. Elective total joint replacement surgery. 2023. Available from: www.ha.org.hk/visitor/ha_visitor_index.asp?Parent_ID=214172&Content_ID=221223#. Accessed 26 Feb 2024.

5. GBD 2021 Osteoarthritis Collaborators. Global, regional, and national burden of osteoarthritis, 1990-2020 and projections to 2050: a systematic analysis for the Global Burden of Disease Study 2021. Lancet Rheumatol 2023;5:e508-22. Crossref

6. Yau WP, Chiu CK, Cheung S, Cheung A. The Hong Kong College of Orthopaedic Surgeons position statement in management of osteoarthritis of knee. Available from: https://www.hkcos.org.hk/Position_Statement/HKCOS_Position_Statement_in_Management_of_Osteoarthritis_of_Knee.pdf. Accessed 26 Feb 2024.

7. American Academy of Orthopaedic Surgeons. Management of osteoarthritis of the knee (nonarthroplasty). Evidence-based clinical practice guideline. Available from: https://www.aaos.org/globalassets/quality-and-practice-resources/osteoarthritis-of-the-knee/oak3cpg.pdf. Accessed 26 Feb 2024.

8. McAlindon TE, Bannuru RR, Sullivan MC, et al. OARSI guidelines for the non-surgical management of knee osteoarthritis. Osteoarthritis Cartilage 2014;22:363-88. Crossref

9. Oxford Centre for Evidence-Based Medicine. OCEBM levels of evidence. Available from: http://www.cebm.ox.ac.uk/resources/levels-of-evidence/ocebm-levels-of-evidence. Accessed 26 Feb 2024.

10. The Hong Kong College of Orthopaedic Surgeons. Statistics. Number of Fellows. Available from: https://www.hkcos.org.hk/aboutus_statistics.asp. Accessed 16 Sep 2024.

11. Driban JB, Harkey MS, Barbe MF, et al. Risk factors and the natural history of accelerated knee osteoarthritis: a narrative review. BMC Musculoskeletal Disord 2020;21:332. Crossref

12. Glyn-Jones S, Palmer AJ, Agricola R, et al. Osteoarthritis. Lancet 2015;386:376-87. Crossref

13. Gwilym SE, Pollard TC, Carr AJ. Understanding pain in osteoarthritis. J Bone Joint Surg Br 2008;90:280-7. Crossref

14. National Institute for Health and Care Excellence. Osteoarthritis in over 16s: diagnosis and management. Available from: http://www.nice.org.uk/guidance/ng226. Accessed 26 Feb 2024.

15. Lane NE, Brandt K, Hawker G, et al. OARSI-FDA initiative: defining the disease state of osteoarthritis. Osteoarthritis Cartilage 2011;19:478-82. Crossref

16. Courties A, Kouki I, Soliman N, Mathieu S, Sellam J. Osteoarthritis year in review 2024: epidemiology and therapy. Osteoarthritis Cartilage 2024;32:1397-404. Crossref

17. Jenei-Lanzl Z, Zaucke F. Osteoarthritis year in review 2024: biology. Osteoarthritis Cartilage 2025;33:58-66. Crossref

18. Matsuda M, Huh Y, Ji RR. Roles of inflammation, neurogenic inflammation, and neuroinflammation in pain. J Anesth 2019;33:131-9. Crossref

19. Cohen SP, Vase L, Hooten WM. Chronic pain: an update on burden, best practices, and new advances. Lancet 2021;397:2082-97. Crossref

20. Saffari M, Emami Meybodi MK, Sanaeinasab H, Karami A, Pakpour AH, Koenig HG. A theory of planned behavior–based intervention to improve quality of life in patients with knee/hip osteoarthritis: a randomized controlled trial. Clin Rheumatol 2018;37:2505-15. Crossref

21. Fransen M, McConnell S, Harmer AR, Van der Esch M, Simic M, Bennell KL. Exercise for osteoarthritis of the knee. Cochrane Database Syst Rev 2015;2015:CD004376. Crossref

22. Silva LE, Valim V, Pessanha AP, et al. Hydrotherapy versus conventional land-based exercise for the management of patients with osteoarthritis of the knee: a randomized clinical trial. Phys Ther 2008;88:12-21. Crossref

23. Chen L, Yao F, Wang T, et al. Horizontal fissuring at the osteochondral interface: a novel and unique pathological feature in patients with obesity-related osteoarthritis. Ann Rheum Dis 2020;79:811-8. Crossref

24. Courties A, Gualillo O, Berenbaum F, Sellam J. Metabolic stress-induced joint inflammation and osteoarthritis. Osteoarthritis Cartilage 2015;23:1955-65. Crossref

25. Puenpatom RA, Victor TW. Increased prevalence of metabolic syndrome in individuals with osteoarthritis: an analysis of NHANES III data. Postgrad Med 2009;121:9-20. Crossref

26. Christensen R, Bartels EM, Astrup A, Bliddal H. Effect of weight reduction in obese patients diagnosed with knee osteoarthritis: a systematic review and meta-analysis. Ann Rheum Dis 2007;66:433-9. Crossref

27. Bliddal H, Leeds AR, Stigsgaard L, Astrup A, Christensen R. Weight loss as treatment for knee osteoarthritis in obese patients: 1-year results from a randomised controlled trial. Ann Rheum Dis 2011;70:1798-803. Crossref

28. Jones A, Silva PG, Silva AC, et al. Impact of cane use on pain, function, general health and energy expenditure during gait in patients with knee osteoarthritis: a randomised controlled trial. Ann Rheum Dis 2012;71:172-9. Crossref

29. Somers TJ, Blumenthal JA, Guilak F, et al. Pain coping skills and lifestyle behavioural weight management in patients with knee osteoarthritis: a randomized controlled study. Pain 2012;153:1199-209. Crossref

30. Hurley MV, Walsh NE, Mitchell HL, et al. Clinical effectiveness of a rehabilitation program integrating exercise, self-management, and active coping strategies for chronic knee pain: a cluster randomized trial. Arthritis Rheum 2007;57:1211-9. Crossref

31. Helminen EE, Sinikallio SH, Valjakka AL, Väisänen-Rouvali RH, Arokoski JP. Effectiveness of a cognitive-behaviour group intervention for knee osteoarthritis pain: a randomized controlled trial. Clin Rehabil 2015;29:868-81. Crossref

32. Schnitzer TJ, Kivitz A, Frayssinet H, Duquesroix B. Efficacy and safety of naproxcinod in the treatment of patients with osteoarthritis of the knee: a 13-week prospective, randomized, multicenter study. Osteoarthritis Cartilage 2010;18:629-39. Crossref

33. Bannuru RR, Osani MC, Vaysbrot EE, et al. OARSI guidelines for the non-surgical management of knee, hip, and polyarticular osteoarthritis. Osteoarthritis Cartilage 2019;27:1578-89. Crossref

34. Kolasinski SL, Neogi T, Hochberg MC, et al. 2019 American College of Rheumatology/Arthritis Foundation guideline for the management of osteoarthritis of the hand, hip, and knee. Arthritis Rheumatol 2020;72:220-3. Crossref

35. Dehghan M, Asgharian S, Khalesi E, Ahmadi A, Lorigooini Z. Comparative study of the effect of Thymus daenensis gel 5% and diclofenac in patients with knee osteoarthritis. Biomedicine (Taipei) 2019;9:9. Crossref

36. Prior MJ, Harrison DD, Frustaci ME. A randomized, double-blind, placebo-controlled 12-week trial of acetaminophen extended release for the treatment of signs and symptoms of osteoarthritis. Curr Med Res Opin 2014;30:2377-87. Crossref

37. Heath EL, Ackerman IN, Cashman K, Lorimer M, Graves SE, Harris IA. Patient-reported outcomes after hip and knee arthroplasty: results from a large national registry. Bone Jt Open 2021;2:422-32. Crossref

38. van Outeren MV, Waarsing JH, Brouwer RW, Verhaar JA, Reijman M, Bierma-Zeinstra SM. Is a high tibial osteotomy (HTO) superior to non-surgical treatment in patients with varus malaligned medial knee osteoarthritis (OA)? A propensity matched study using 2 randomized controlled trial (RCT) datasets. Osteoarthritis Cartilage 2017;25:1988-93. Crossref

39. Wilson HA, Middleton R, Abram SG, et al. Patient relevant outcomes of unicompartmental versus total knee replacement: systematic review and meta-analysis. BMJ 2019;364:I352. Crossref

The coronavirus disease 2019 (COVID-19) pandemic has had a broad public health impact on global populations. Pandemic-control measures (eg, social distancing regulations and changes in hospital services) have affected both the general population and individuals with chronic diseases (including cancer survivors).1 A recent meta-analysis found that 53.9% of cancer patients (n=27 590) experienced high levels of distress during the COVID-19 pandemic2; breast cancer survivors (BCS) reported highest prevalence of post-traumatic stress symptoms (52.3%) among all groups of cancer patients.2 In Hong Kong, 40% of cancer survivors reported feeling anxious and depressed during COVID-19.3 However, factors associated with psychological distress among local BCS during the pandemic have been understudied.

After completion of active treatment, BCS require a range of supportive cancer care services for rehabilitation. To prioritise resources for managing COVID-19, some oncology services were postponed in Hong Kong.4 Among a heterogeneous sample of cancer survivors in Hong Kong,3 less than10% reported that the COVID-19 pandemic had affected their hospital treatments or follow-ups. Despite this low prevalence, the potential negative impacts of such disruptions on cancer survivors’ well-being should not be ignored. A systematic review found that delays or changes in treatment plans were associated with high levels of psychological distress, above and beyond the contributions of other socio-demographic factors.5 Similarly, BCS in the United Kingdom who experienced disrupted oncology services reported worse emotional well-being.6 In China, a 3-week treatment delay was significantly associated with increased psychological symptoms among BCS.7 Based on these findings, we speculated that cancer treatment and supportive care service disruptions would be associated with greater psychological distress among BCS in Hong Kong.

Given that cancer survivors are more aware of the risks of infection compared with the general population,8 their emotional reactions and perceptions towards COVID-19 also contribute to their well-being. Due to the rapidly changing pandemic situations caused by different variants of the COVID-19 virus, cancer survivors tend to experience fear of contracting COVID-19 and express concerns about the severity of its negative health impacts on cancer prognosis.9 10 Based on a review of 51 studies (19.5% conducted in Asia), COVID-19–related fear and worries were associated with psychological distress among cancer survivors.11 However, a recent study in Hong Kong showed that 49.7% of cancer survivors did not feel worried about contracting COVID-19,3 and they did not consider themselves to experience more negative consequences of contracting COVID-19 compared with the general population.12 Whether COVID-19–related fear and risk perceptions are associated with psychological distress among BCS in Hong Kong has yet to be explored.

The COVID-19 pandemic has led to unmet supportive care needs (SCNs) among BCS. According to Fitch’s Supportive Care Needs Framework,13 a medical diagnosis affects people’s abilities to meet their own needs across life domains; unmet needs may result in worse adjustment outcomes. It is common for cancer survivors to report physical, psychological, social, and health system–related unmet needs during COVID-19.14 A longitudinal survey of Asian and Asian American cancer survivors revealed a significant increase in psychological and healthcare access needs during COVID-19.15 In Hong Kong, higher levels of unmet SCNs were identified in the health system/information, psychological, and patient care/support domains among cancer survivors during COVID-19.3 In Australia, unmet SCNs were associated with greater psychological distress among haematological and gynaecological cancer survivors.16 However, research examining the associations between unmet SCNs and BCS’ psychological distress has been sparse.

This study examined factors associated with psychological distress among Hong Kong BCS during COVID-19. We hypothesised that different domains of unmet SCNs, disruptions in cancer treatment and supportive cancer care services, increased fear of COVID-19, and a stronger belief that COVID-19 would cause more severe health consequences for BCS (compared with the general population) were associated with greater psychological distress.

Prospective participants were recruited from the Hong Kong Breast Cancer Registry, the most representative monitoring system for BCS in Hong Kong.17 Based on the cancer registry data, those fulfilling the inclusion criteria were invited to participate in a cross-sectional survey. Breast cancer survivors eligible for the study were required to be: aged 18 years or above, diagnosed with stages 0 to III cancer since January 2020, in active treatment, able to read Chinese and communicate in Cantonese, and able to provide informed consent.

Among 946 BCS contacted, 409 were unreachable, 23 were ineligible, and 227 were uninterested in the study. With verbal consent given over the phone, those who were eligible and interested in the study (n=287) received a mail package enclosing a cover letter explaining the study details, a consent form, a questionnaire packet, a stamped return envelope, and a thank-you card. After they had provided consent, participants completed the survey at home. Participants were compensated with supermarket vouchers (worth HK$100) for their time upon returning the completed survey. The study was conducted between June and December 2022. Overall, 209 completed surveys were returned (from 287 sent), yielding a completion rate of 72.8%.

The one-item National Comprehensive Cancer Network Distress Thermometer was used to assess participants’ psychological distress over the past week.18 On an 11-point Likert scale, the thermometer ranged from 0 (no distress) to 10 (extreme distress).19 A higher score indicated a higher level of psychological distress. A cut-off point of ≥4 indicated moderate-to-severe distress.20 The Chinese version of the Distress Thermometer has demonstrated reliability and validity among Chinese cancer patients.21

Participants’ experiences of any postponement or cancellation of various types of cancer treatments (eg, surgery and adjuvant therapies) and supportive cancer care services (eg, psychological counselling and peer support groups) during COVID-19 were measured (no=0, yes=1).

The Chinese version of the 34-item Short-Form Supportive Care Needs Questionnaire was used to measure five domains of SCNs (namely, physical/daily living, psychological needs, patient care and support, sexuality, and health system/information needs) over the past month. On a five-point scale (no need–not applicable, no need–satisfied, low need, moderate need, high need), items were scored using standardised guidelines.22 Higher scores indicated higher levels of unmet SCNs. The scale has shown reliability and validity among Hong Kong cancer survivors.3 22

The Chinese version of the seven-item Fear of COVID-19 Scale was adapted to measure participants’ fear of COVID-19.23 On a five-point scale (1=strongly disagree, 5=strongly agree), a higher mean score indicated greater fear of COVID-19 (eg, “My heart races or palpitates when I think about getting COVID-19”; Cronbach’s α=0.88). The scale has demonstrated reliability and validity in a Chinese general population.23

A single item was developed to measure participants’ perception of the severity of COVID-19 health consequences for BCS (ie, “COVID-19 can cause more severe health consequences in BCS than in the general population”). Responses were recorded on a five-point scale (1=strongly disagree, 5=strongly agree); higher scores indicated greater perceived severity.

Participants self-reported the following characteristics: (1) socio-demographic information; (2) treatment-related variables (surgery undergone, treatments being received or completed, and time since last treatment); and (3) breast cancer–related variables (eg, stage at diagnosis and time since diagnosis).

Descriptive statistics and bivariate correlations among the variables were computed. Multivariable logistic regressions were used to examine associations between independent variables and moderate-to-severe levels of psychological distress (binary-coded), based on the suggested Distress Thermometer cut-off of ≥4. Simple logistic regression analysis was used to assess how individual variables (including socio-demographic/cancer-related variables, cancer treatment/supportive cancer care service disruptions, unmet SCNs, fear of COVID-19, and COVID-19 risk perception) were associated with psychological distress. Odds ratios (ORs) were obtained by separately fitting each variable against psychological distress.24 Significant variables in the simple analyses were then entered into a multivariable logistic regression model using the enter method. These analyses were performed using SPSS (Windows version 26.0; IBM Corp, Armonk [NY], US). P values <0.05 were considered statistically significant.

Based on prior studies regarding BCS’ psychological distress,24 25 we assumed a similar prevalence of 30% for moderate-to-severe levels of psychological distress in our target population. With α=0.05 (two-tailed) and a statistical power of 80%, a sample size of 203 would be sufficient to detect an OR of 1.56 for the key independent variables (G^*Power version 3.1.2).26 The current sample size (n=209) was sufficient to detect the expected effect sizes with adequate statistical power.

Among the 209 participants, 69.3% were aged ≤60 years. Regarding cancer-related characteristics, 4.8%, 28.7%, 41.1%, and 25.4% were diagnosed with Stage 0, Stage I, Stage II, and Stage III breast cancer, respectively. Most participants (94.3%) had undergone breast cancer surgery. Most had also received chemotherapy (72.7%) and radiotherapy (76.1%). The average time since diagnosis was 16.6 months (standard deviation=8.00). During COVID-19, 32.1% of participants had a prior diagnosis of COVID-19; 20.1% experienced cancer treatment disruptions (eg, surgery/adjuvant therapies); and 49.3% experienced supportive cancer care service disruptions (eg, psychological counselling, patient support groups). Based on a Distress Thermometer score ≥4, 50.2% of participants reported a moderate-to-severe level of psychological distress (Table 1).

Based on the correlation analysis results (Table 2), participants who had experienced supportive cancer care service disruption, perceived greater fear of COVID-19, and held a stronger belief that COVID-19 causes more severe health consequences in BCS tended to report increased psychological distress (sample correlation coefficients=0.16-0.34; all P<0.01). All five domains of unmet SCNs were associated with increased psychological distress (sample correlation coefficients=0.33-0.64; all P<0.001).

In the simple logistic regression analyses (Table 3), no background variables showed significant associations with psychological distress; accordingly, adjustments for those variables were not included in the final multivariable regression model. On the other hand, all domains of unmet SCNs (ORs=1.03-1.07; all P<0.001), cancer supportive care service disruption (OR=1.75; P=0.05), fear of COVID-19 (OR=2.11; P=0.001), and the perception that COVID-19 causes more severe health consequences in BCS (OR=1.72; P<0.001) were associated with moderate-to-severe psychological distress (Table 4). The multivariable logistic regression results indicated that only unmet physical needs (OR=1.03; P=0.002), unmet psychological needs (OR=1.06; P<0.001), and perceived severity of COVID-19–related health consequences in BCS (OR=1.67; P=0.02) were associated with moderate-to-severe psychological distress (Table 4).

At least 50% of BCS in Hong Kong experienced a moderate-to-severe level of distress during COVID-19. This prevalence was comparable to that of gynaecological cancer survivors in Turkey,27 but lower than that of sarcoma patients in Italy.28 Discrepancies in prevalence might be attributed to varied pandemic situations across regions, including differences in survey periods, cancer types, and specific pandemic-control measures. Future research could investigate how these factors jointly contribute to BCS’ psychological distress. Among the studied variables, multivariable logistic regression analysis revealed that higher levels of unmet SCNs in physical and psychological domains, along with a stronger belief that COVID-19 could cause more severe health consequences in BCS, were associated with greater psychological distress among BCS in Hong Kong.

We found that 20.1% and 49.3% of BCS experienced cancer treatment and supportive cancer care service disruptions, respectively, during COVID-19. Only the disruption of supportive cancer care services (but not cancer treatments) demonstrated a significant univariate association with moderate-to-severe psychological distress. Previously, cancer care service disruptions during COVID-19 were associated with worse psychological outcomes among BCS in the United Kingdom,6 Canada,29 and Ireland.30 Given that disruptions of supportive cancer care services (but not cancer treatments) were positively associated with unmet SCNs, the absence of timely supportive care might make coping with and living with cancer particularly difficult during COVID-19.11 However, supportive cancer care service disruption was no longer significant in multivariable logistic regression analyses when other independent variables were considered. This finding implies that more proximal factors related to BCS’ daily lives and challenges (eg, different domains of unmet SCNs) had relatively greater prominence in explaining psychological distress among BCS in Hong Kong during COVID-19.

Breast cancer survivors in Hong Kong reported moderate levels of unmet SCNs across different domains, with the highest score in the health system/information domain and the lowest score in the sexuality domain (Table 5). These levels of unmet SCNs were comparable to those reported by other local cancer survivors in 2021.3 Our findings showed that all five domains of unmet SCNs were associated with moderate-to-severe psychological distress, but only unmet SCNs in the psychological and physical/daily living domains constituted significant contributors in the final multivariable logistic regression model (Table 4). These findings were similar to a study conducted among survivors of mixed cancer types (56.7% diagnosed within the previous year) in Turkey.31 According to that study,31 conducted during COVID-19, all domains of unmet SCNs were correlated with depression/anxiety, but only unmet SCNs in the psychological and physical/daily living domains independently contributed to depression/anxiety in the multivariable analysis. In contrast, among cancer survivors in Jordan (77% diagnosed ≥6 years prior), all domains of unmet SCNs during COVID-19 were independently associated with quality of life.32 Given that our participants were recently diagnosed BCS (with an average time since diagnosis of 16.6 months; all diagnosed after the COVID-19 pandemic began), they were still in the process of coping with the reality of the cancer diagnosis, the discomfort and side-effects of treatment, uncertainties about the future, and potential cancer recurrences. Considering the time since diagnosis, the relative contributions of SCNs in the psychological and physical/daily living domains to psychological distress were particularly strong among recently diagnosed BCS during COVID-19.

Fear of COVID-19 was associated with greater psychological distress only in the simple analysis, but not in the multivariable regression model. Previously, fear of COVID-19 was associated with greater psychological distress among cancer survivors in the US33 and the general population in Hong Kong34 during earlier phases of the COVID-19 pandemic in 2020 to 2021. Due to Hong Kong’s unique experiences in successfully managing prior pandemics (eg, the severe acute respiratory syndrome and the H1N1 pandemics),35 pandemic fatigue (ie, a state of emotional and physical exhaustion resulting from prolonged anti-pandemic measures) was observed during the fourth and fifth waves of the pandemic in Hong Kong (2021-2022).36 Such fatigue was reflected in the lower levels of fear of COVID-19 (ie, affective and physiological states of anxiety and fear towards COVID-19) among our sample surveyed in 2022, compared with the general population surveyed in early 2021,34 using the same measurement. This finding might explain why the contribution of fear of COVID-19 to psychological distress among local BCS was weaker than expected.

Conversely, we found that the perceived severity of the health consequences of COVID-19 for BCS was a stronger contributor to psychological distress than fear of COVID-19. Cancer and its treatments (eg, chemotherapy) can weaken patients’ immune systems, and it is common for BCS to believe that being immunocompromised might lead to more severe health consequences if they contract COVID-19.1 Risk perception has been associated with coping behaviours. For example, a recent study found that risk perception about COVID-19 was a stronger contributor to information-seeking behaviour among the general population in Hong Kong than among their counterparts in China and Taiwan.37 We expected that this phenomenon would also be apparent among Hong Kong BCS. However, health information about COVID-19 may not always be tailored for cancer survivors or effectively communicated through local mass media,1 which could be associated with psychological distress among BCS. To alleviate such COVID-19–related risk perceptions, we recommend that health organisations tailor health information and provide counselling for cancer survivors through alternative platforms (eg, social media and online forums).

This study had several limitations. First, given its cross-sectional design, it could not establish causal relationships among the variables. Cancer survivors’ risk perceptions about COVID-19 and unmet SCNs are likely to change throughout the course of their cancer journey. Future studies could utilise longitudinal designs to better understand the temporal relationships among variables and psychological distress. Second, although the Hong Kong Breast Cancer Registry is the most comprehensive registry for BCS in Hong Kong, it does not cover the entire BCS population due to its voluntary enrolment system. The generalisability of the findings to BCS in other countries with different healthcare systems and pandemic situations should be interpreted with caution. Third, other important independent variables might contribute to BCS’ psychological distress. Studies have revealed that additional daily COVID-19 stressors (eg, increased responsibilities at home and difficulties obtaining daily necessities) and coping strategies (eg, catastrophising) may play key roles in explaining psychological distress among cancer survivors.12 29 The inclusion of such variables could further improve the explanatory power of the regression model. Fourth, to reduce participant burden, we measured risk perceptions related to COVID-19 using a self-developed item. Specifically developed items are commonly used as predictors of psychological outcomes to capture nuances in the local COVID-19 context.38 However, researchers are encouraged to confirm our findings using fully validated instruments for the measurement of COVID-19 risk perceptions.

This study highlights the importance of addressing BCS’ unmet SCNs in the physical/daily living and psychological domains, as well as their risk perceptions of COVID-19, in relation to psychological distress during the pandemic. To address physical/daily living needs, survivors might need to engage in self-monitoring of health (eg, reporting symptoms and metrics to healthcare providers through patient portals). Psychological well-being should be regularly monitored, and communication between providers and survivors should be maintained through virtual means. In addition to information about cancer symptom management, survivors should be provided with accessible mental health services that can support them in coping with the emotional impacts of their diagnosis and treatment.39 Regarding COVID-19 risk perception, it may be beneficial to offer accurate and up-to-date educational materials explaining BCS’ risks associated with COVID-19 and how they can protect themselves. Research suggests that telehealth can empower survivors and provide strategies for coping during unprecedented times. A recent study in Iran indicated that a tele-nursing intervention—including supportive telephone calls with explanations about cancer, treatment side-effects, symptom management, and self-care—reduced unmet SCNs among Iranian cancer survivors undergoing chemotherapy.40 Researchers should explore the applicability of such service models in Hong Kong and other regions.

Half of BCS in Hong Kong experienced a moderate-to-severe level of psychological distress during COVID-19. Efforts to address unmet SCNs in the physical/daily living and psychological domains, manage risk perceptions regarding health consequences of COVID-19, and provide supportive cancer care services through alternative modes might help alleviate psychological distress among BCS in future pandemic situations.

Concept or design: All authors.
Acquisition of data: NCY Yeung, STY Lau.
Analysis or interpretation of data: NCY Yeung, STY Lau.
Drafting of the manuscript: NCY Yeung, STY Lau.
Critical revision of the manuscript for important intellectual content: NCY Yeung.

All authors had full access to the data, contributed to the study, approved the final version for publication, and take responsibility for its accuracy and integrity.

All authors have disclosed no conflicts of interest.

This research was supported by the Health and Medical Research Fund of the Health Bureau, Hong Kong SAR Government (Ref No.: 18190061). The funder had no role in the study design, data collection/analysis/interpretation, or manuscript preparation.

This research was approved by the Joint Chinese University of Hong Kong–New Territories East Cluster Clinical Research Ethics Committee, Hong Kong (Ref No.: 2021.286) and the Hong Kong Breast Cancer Foundation. Informed consent was obtained from all individual participants included in the study.

1. Huang J, Wang HH, Zheng ZJ, Wong MC. Impact of the COVID-19 pandemic on cancer care. Hong Kong Med J 2022;28:427-9. Crossref

2. Zhang H, Xu H, Zhang Z, Zhang Q. Efficacy of virtual reality–based interventions for patients with breast cancer symptom and rehabilitation management: a systematic review and meta-analysis. BMJ Open 2022;12:e051808. Crossref

3. Cheung DS, Takemura N, Lui HY, et al. A cross-sectional study on the unmet supportive care needs of cancer patients under the COVID-19 pandemic. Cancer Care Res Online 2022;2:e028. Crossref

4. Koczwara B. Cancer survivorship care at the time of the COVID-19 pandemic. Med J Aust 2020;213:107-8.e1. Crossref

5. Prabani KI, Damayanthi HD. Quality of life, anxiety, depression and psychological distress in patients with cancer during the COVID 19 pandemic: a systematic review. Intl Health Trends Perspect 2022;2:51-66. Crossref

6. Swainston J, Chapman B, Grunfeld EA, Derakshan N. COVID-19 lockdown and its adverse impact on psychological health in breast cancer. Front Psychol 2020;11:2033. Crossref

7. Wang Y, Yang Y, Yan C, et al. COVID-induced 3 weeks’ treatment delay may exacerbate breast cancer patient’s psychological symptoms. Front Psychol 2022;13:1003016. Crossref

8. Slivjak ET, Fishbein JN, Nealis M, Schmiege SJ, Arch JJ. Cancer survivors’ perceived vulnerability to COVID-19 and impacts on cognitive, affective, and behavioral responses to the pandemic. J Psychosoc Oncol 2021;39:366-84. Crossref

9. Glidden C, Howden K, Romanescu RG, et al. Psychological distress and experiences of adolescents and young adults with cancer during the COVID-19 pandemic: a cross-sectional survey. Psychooncology 2022;31:631-40. Crossref

10. Shay LA, Allicock M, Li A. “Every day is just kind of weighing my options.” Perspectives of young adult cancer survivors dealing with the uncertainty of the COVID-19 global pandemic. J Cancer Surviv 2022;16:760-70. Crossref

11. Muls A, Georgopoulou S, Hainsworth E, et al. The psychosocial and emotional experiences of cancer patients during the COVID-19 pandemic: a systematic review. Semin Oncol 2022;49:371-82. Crossref

12. Ng DW, Chan FH, Barry TJ, et al. Psychological distress during the 2019 coronavirus disease (COVID-19) pandemic among cancer survivors and healthy controls. Psychooncology 2020;29:1380-3. Crossref

13. Fitch MI. Supportive Care Framework [in English, French]. Can Oncol Nurs J 2008;18:6-24. Crossref

14. Legge H, Toohey K, Kavanagh PS, Paterson C. The unmet supportive care needs of people affected by cancer during the COVID-19 pandemic: an integrative review. J Cancer Surviv 2023;17:1036-56. Crossref

15. Wang K, Ma C, Li FM, et al. Patient-reported supportive care needs among Asian American cancer patients. Support Care Cancer 2022;30:9163-70. Crossref

16. Zomerdijk N, Jongenelis M, Short CE, Smith A, Turner J, Huntley K. Prevalence and correlates of psychological distress, unmet supportive care needs, and fear of cancer recurrence among haematological cancer patients during the COVID-19 pandemic. Support Care Cancer 2021;29:7755-64. Crossref

17. Hong Kong Breast Cancer Foundation. Hong Kong Breast Cancer Registry Report No. 11. 2019. Available from: https://www.hkbcf.org/en/our_research/main/468/. Accessed 26 May 2023.

18. National Comprehensive Cancer Network: Distress Thermometer Tool Translations. Available from: https://www.nccn.org/global/what-we-do/distress-thermometer-tool-translations. Accessed 10 Jul 2023.

19. Ownby KK. Use of the Distress Thermometer in clinical practice. J Adv Pract Oncol 2019;10:175-9. Crossref

20. Wood DE. National Comprehensive Cancer Network (NCCN) clinical practice guidelines for lung cancer screening. Thorac Surg Clin 2015;25:185-97. Crossref

21. Thapa S, Sun H, Pokhrel G, Wang B, Dahal S, Yu S. Performance of Distress Thermometer and associated factors of psychological distress among Chinese cancer patients. J Oncol 2020;2020:3293589. Crossref

22. Au A, Lam W, Tsang J, et al. Supportive care needs in Hong Kong Chinese women confronting advanced breast cancer. Psychooncology 2013;22:1144-51. Crossref

23. Chi X, Chen S, Chen Y, et al. Psychometric evaluation of the Fear of COVID-19 Scale among Chinese population. Int J Ment Health Addict 2022;20:1273-88. Crossref

24. Vanni G, Materazzo M, Santori F, et al. The effect of coronavirus (COVID-19) on breast cancer teamwork: a multicentric survey. In Vivo 2020;34(3 Suppl):1685-94. Crossref

25. Pang L, Yao S, Li W, Jing Y, Yin X, Cheng H. Impact of the CALM intervention on breast cancer patients during the COVID-19 pandemic. Support Care Cancer 2023;31:121. Crossref

26. Faul F, Erdfelder E, Buchner A, Lang AG. Statistical power analyses using G*Power 3.1: tests for correlation and regression analyses. Behav Res Methods 2009;41:1149-60. Crossref

27. Yavuz AD, Dolanbay M, Akyüz Çim EF, Dişli Gürler A, Cündübey CF. Analysis of distress in patients with gynecological cancers during the COVID-19 pandemic: a telephone survey. J Clin Obstet Gynecol 2022;32:93-9. Crossref

28. Onesti CE, Vari S, Minghelli D, et al. Quality of life and emotional distress in sarcoma patients diagnosed during COVID-19 pandemic: a supplementary analysis from the SarCorD study. Front Psychol 2023;14:1078992. Crossref

29. Massicotte V, Ivers H, Savard J. COVID-19 pandemic stressors and psychological symptoms in breast cancer patients. Curr Oncol 2021;28:294-300. Crossref

30. Myers C, Bennett K, Kelly C, et al. Impact of COVID-19 on health care and quality of life in women with breast cancer. JNCI Cancer Spectr 2023;7:pkad033. Crossref

31. Erdoğan Yüce G, Döner A, Muz G. Psychological distress and its association with unmet needs and symptom burden in outpatient cancer patients: a cross-sectional study. Semin Oncol Nurs 2021;37:151214. Crossref

32. Al-Omari A, Al-Rawashdeh N, Damsees R, et al. Supportive care needs assessment for cancer survivors at a comprehensive cancer center in the Middle East: mending the gap. Cancers (Basel) 2022;14:1002. Crossref

33. Caston NE, Lawhon VM, Smith KL, et al. Examining the association among fear of COVID-19, psychological distress, and delays in cancer care. Cancer Med 2021;10:8854-65. Crossref

34. Chair SY, Chien WT, Liu T, et al. Psychological distress, fear and coping strategies among Hong Kong people during the COVID-19 pandemic. Curr Psychol 2023;42:2538-57. Crossref

35. Matus K, Sharif N, Li A, Cai Z, Lee WH, Song M. From SARS to COVID-19: the role of experience and experts in Hong Kong’s initial policy response to an emerging pandemic. Humanit Soc Sci Commun 2023;10:9. Crossref

36. Leung HT, Gong W, Sit SM, et al. COVID-19 pandemic fatigue and its sociodemographic and psycho-behavioral correlates: a population-based cross-sectional study in Hong Kong. Sci Rep 2022;12:16114. Crossref

37. Liu R, Huang YC, Sun J. The media-mediated model of information seeking behavior: a proposed framework in the Chinese culture during the COVID-19 pandemic. Health Commun 2024;39:3468-79. Crossref

38. Yeung NC, Tang JL, Hui KH, Lau ST, Cheung AW, Wong EL. “The light after the storm”: psychosocial correlates of adversarial growth among nurses in Hong Kong amid the fifth wave of the COVID-19 pandemic. Psychol Trauma 2024;16:989-98. Crossref

39. Thompson CA, Overholser LS, Hébert JR, Risendal BC, Morrato EH, Wheeler SB. Addressing cancer survivorship care under COVID-19: perspectives from the cancer prevention and control research network. Am J Prev Med 2021;60:732-6. Crossref

40. Ebrahimabadi M, Rafiei F, Nejat N. Can tele-nursing affect the supportive care needs of patients with cancer undergoing chemotherapy? A randomized controlled trial follow-up study. Support Care Cancer 2021;29:5865-72. Crossref

Over the past few decades, the widespread adoption of robotic surgical systems has revolutionised surgical management, particularly for radical prostatectomy.1 Enhanced visualisation, superior dexterity, and tremor filtration enable surgeons to maintain precision in the deep pelvis with better ergonomics. These advantages contribute to lower rates of positive surgical margins and biochemical recurrence, thereby reducing the need for salvage therapy.2 Evidence also shows that robotic surgery can significantly reduce postoperative complications and shorten the time to regain continence and potency.3

For the past two decades, the global market for robotic surgery has been dominated by the da Vinci Surgical System (Intuitive Surgical, Sunnyvale [CA], US). However, worldwide adoption of robotic surgery remains limited by the high costs associated with device acquisition, maintenance, and disposables. Consequently, access disparities persist worldwide, particularly exacerbating inequities in surgical care4 in low- and middle-income countries.5 Surgeons and trainees are also deprived of opportunities to acquire robotic skills, creating a bottleneck in the clinical application of advanced surgical technologies. To address this problem, a novel, high-performing yet affordable robotic surgical system is needed to expand access to the highest standards of surgical care.

The Sentire Surgical System (C1000) is a novel robotic surgical system developed by Cornerstone Robotics Limited (Hong Kong, China). It is the first locally developed surgical robot designed for multispecialty use. The system consists of three interconnected components: a Patient-Side Robot (PSR), a Surgeon Console, and a Vision Cart. The PSR has four robotic arms, on which an endoscope and up to three robotic surgical instruments can be mounted. The endoscope provides a high-resolution, three-dimensional image for the console surgeon. An array of instruments, including graspers, needle drivers, clip appliers, and scissors, can be installed depending on the operational requirements. The Surgeon Console includes two hand controls and a set of foot pedals, enabling surgeons to operate the surgical instruments and endoscope on the PSR, while applying energy through the surgical instruments. The Vision Cart includes a touchscreen display showing the endoscopic view for the assistant surgeon. It houses the energy source for monopolar and bipolar instruments, as well as the light and camera source for the endoscope.

This study aimed to evaluate the clinical safety and efficacy of the Sentire Surgical System by reporting outcomes from the first clinical trial for radical prostatectomy, as part of a multispecialty clinical study. Comparable safety and patient outcomes will provide supporting evidence for the continued development of this robotic surgical technology, facilitating its implementation and evaluation in further clinical trials.

This was a prospective, single-centre, single-arm study aligned with Stage 1 (Innovation) of the IDEAL (Innovation, Development, Exploration, Assessment, Long-term Study) framework.6 The study formed part of a multi-speciality clinical investigation to evaluate the safety and efficacy of the Sentire Surgical System for robot-assisted colorectal, upper gastrointestinal, and urological surgery (radical prostatectomy).

From August 2022 to September 2023, 20 adult men with clinically localised prostate cancer and planned radical prostatectomy were recruited. The inclusion criteria were: (1) age between 50 and 80 years; (2) clinical diagnosis of non–metastatic prostate cancer; (3) body mass index <35 kg/m²; (4) deemed suitable for minimally invasive treatment; and (5) provision of informed consent. Exclusion criteria were: (1) contraindication to general anaesthesia; (2) prior history of prostatic surgery; (3) untreated active infection; (4) uncorrected coagulopathy; (5) presence of other malignancies or distant metastases; and (6) membership in a vulnerable population.

All 20 robot-assisted radical prostatectomies were performed by four practising urological surgeons, each with extensive experience (>50 prior cases as chief surgeon). All participating surgeons had been trained in the use of the system and had prior experience performing procedures with the system on cadaveric and live porcine models.

Patients were placed under general anaesthesia in the lithotomy position. A total of four robotic ports were used. First, a 10-mm endoscope port was placed supraumbilically through an open approach. After pneumoperitoneum was established, three additional 8-mm robotic ports were inserted: one on either side of the supraumbilical port, adjacent to the left mid-clavicular line at approximately the vertical level of the umbilicus, and one along the right or left anterior axillary line, about 2 cm above the anterior iliac spine. A 12-mm assistant port was inserted opposite the most lateral robotic port, mirrored across the midline. Finally, a 5-mm assistant port was inserted between the midline and either the right or left mid-clavicular line (on the same side as the 12-mm assistant port) for suction and retraction (Fig 1).

Following port placement, the patient was placed in the Trendelenburg position. The PSR of the Sentire Surgical System, which consists of four robotic arms extending from a central column, was positioned between the patient’s legs. Figure 2 shows photos of the operating room during radical prostatectomy.

For most radical prostatectomy procedures, a 0° endoscope was used throughout. One surgeon alternated among the 0°, 30° up (for initial bladder detachment), and 30° down endoscopes (for bladder neck dissection) according to their preference. Bipolar Maryland Forceps (Cornerstone Robotics Limited, Hong Kong SAR, China) were used in the left hand, Monopolar Curved Scissors (Cornerstone Robotics Limited, Hong Kong SAR, China) in the right hand, and ExtraGrasp forceps (Cornerstone Robotics Limited, Hong Kong SAR, China) on the fourth robotic arm for traction. Left-hand and right-hand instruments were exchanged for the Large Needle Driver during vesicourethral anastomosis and other suturing tasks. The right-hand instrument could also be replaced with the Large Clip Applier for vessel ligation.

Each operation was performed using the standard transperitoneal anterior approach. Briefly, bladder dissection was carried out in the areolar tissue plane between the peritoneum and transversalis fascia after the peritoneum had been incised at the lateral borders of the urinary bladder. Dissection continued caudally to develop the retropubic space. Preprostatic fat was removed for enhanced anatomical localisation. Dissection of the endopelvic fascia was then performed. The dorsal venous complex was ligated according to surgeon preference. Bladder neck dissection followed, then dissection of the vas deferens and seminal vesicles. The posterior plane of the prostate was then dissected, continuing to the prostatic apex. Lateral dissection was performed next, with the degree of nerve sparing determined by preoperative erectile function, tumour location, and tumour grade. Apical dissection was performed and the urethra was then divided.

Haemostasis was secured by further suturing of the dorsal venous complex, if not previously completed. If indicated, pelvic lymph node dissection was performed at this stage. Posterior Rocco’s stitch reconstruction was carried out, followed by vesicourethral anastomosis with continuous sutures. After confirming watertightness, a pelvic drain was placed via the most lateral robotic port. The robot was then undocked and specimens were removed in a specimen bag through an extended camera port incision. Wounds were closed using standard techniques.

Primary endpoints included the conversion rate and the incidence of perioperative complications within 30 days. Conversion was defined as an emergent change to a conventional laparoscopic or open approach. Intraoperative events and postoperative complications during the hospital stay and within 30 days of discharge were recorded. The severity of complications was graded according to the Clavien–Dindo Classification. Whether complications were anticipated was determined based on their consistency with the current investigational plan or consent form. In the event of postoperative complications, the relationship to the Sentire Surgical System, the surgical procedure, and the patient’s underlying condition was documented, along with actions taken and outcomes.

Secondary endpoints included perioperative and pathological outcomes. Perioperative outcomes included total operative time, docking time, total console time, estimated blood loss, time to resume regular activity, length of hospital stay, time of drain and removal, time of urethral catheter removal, visual analogue scale (VAS) pain scores at 14 and 30 days, and use of pads for urinary incontinence at 14 days and 30 days. Pathological outcomes included surgical margins, lymph node metastasis, tumour involvement, pathological tumour stage, number of lymph nodes harvested, and postoperative prostate-specific antigen (PSA) level.

Data are presented in a descriptive manner. Continuous variables are reported as means with standard deviations (SDs) or medians with interquartile ranges, whereas categorical variables are presented as percentages. All analyses were performed using SPSS (Windows version 25.0; IBM Corp, Armonk [NY], US).

Twenty patients were enrolled in the trial, and all successfully underwent robot-assisted radical prostatectomy using the Sentire Surgical System. The mean age was 68.5 years (SD=4.1) and the mean body mass index was 24.2 kg/m² (SD=2.8). Two patients had a history of abdominal surgery: Case 3 had undergone open appendicectomy, and Case 11 had undergone laparoscopic cholecystectomy. Seventeen patients underwent radical prostatectomy only, while three also underwent lymphadenectomy. The mean prostate volume was 40.7 cc (SD=17.8). Nine patients had a Gleason score of 6, nine had a score of 7, and two had a score of 8. Five patients were classified as low-risk, 11 as intermediate-risk, and four as high-risk. Two patients received neoadjuvant hormonal therapy due to concerns about prolonged waiting times during the coronavirus disease 2019 period; their mean initial PSA level dropped from 16.2 ng/mL to 2.19 ng/mL. The mean preoperative PSA level for patients who did not receive neoadjuvant therapy was 15.2 ng/mL (SD=20.8) [Table 1].

All operations were completed without conversion. There were no intraoperative complications related to the robotic device. Minor surgical complications (Grade I-II) occurred in seven patients, including four cases of wound infection, one case of urinary retention, one case of prolonged anastomotic leakage, one case of bilateral lower limb rash, and one case of left loin pain. Only five of these complications—namely wound infections and prolonged anastomotic leakage—were considered related to the surgery (Table 2).

The mean total operative time (from skin incision to closure) was 184.5 minutes, including a mean docking time of 4.2 minutes and a mean total console time of 149.7 minutes. The median estimated blood loss was 175.0 mL. Postoperatively, the median time to resumption of regular activity was 7.0 days, the median duration of drainage was 2.0 days and the median duration of urethral catheterisation was 7.0 days. The median length of hospital stay was 3.0 days. The median VAS pain score at the 14-day follow-up was 3.0, whereas at the 30-day follow-up it was 0. Pad usage for urinary incontinence was 3.5 at the 14-day follow-up and 2.0 at the 30-day follow-up (Table 3).

Regarding pathological outcomes, 13 of the 20 patients (65%) had negative surgical margins, including the two who received neoadjuvant hormonal therapy. Seventeen patients (85%) had undetectable PSA levels 1 month after surgery, while two other patients (95%) achieved undetectable levels at 3 months. Among the three patients who underwent lymphadenectomy, the mean number of lymph nodes harvested was 9 (range=5-11) [Table 4].

This prospective study reports the results of the first-in-human clinical trial of the Sentire Surgical System—the first locally developed multispecialty surgical robot—used for radical prostatectomy. All 20 procedures were completed successfully, without conversions and device-related intraoperative or postoperative complications, demonstrating that this novel surgical system is safe and effective for robot-assisted radical prostatectomy.

Since the introduction of robotic surgery in Hong Kong in 2005, its use—particularly in urology—has steadily expanded.7 According to the latest Surgical Outcome Monitoring and Improvement Programme report (2022-2023), almost all radical prostatectomies (409 of 410 patients) were performed using robotic surgery.8 Improved outcomes have been a key factor supporting its development.9 However, the high cost of robotic systems considerably limits further expansion and popularisation. With the expiration of certain technological patents, new robotic systems are rapidly emerging worldwide.10 The development of a locally based robotic system in Hong Kong represents an important milestone for the future of robotic surgery in the region.

Perioperative outcomes in this trial were favourable, with a mean total operative time of 184.5 minutes, comparable to reported times in the existing literature on both laparoscopic and robotic radical prostatectomy.11 The short docking time reflected a smooth docking process. Most cases had acceptable blood loss, and no patients required transfusion, again comparable to the literature.11 Despite the higher-than-expected wound infection rate, most infections were minor and managed with simple dressing. No specific cause was identified, and these events are unlikely to be related to the robotic system. Postoperative hospital stay was also short, with satisfactorily low VAS pain scores. Early functional outcomes, specifically pad usage for urinary incontinence, were also satisfactory (Table 3).

Short-term oncological outcomes also support the system’s efficacy, such that 65% of patients achieved a negative surgical margin and 85% and 95% of patients showed no PSA persistence (PSA level ≥0.1 ng/mL) at 30 days and 3 months after surgery, respectively—results that are comparable to the literature.12 In the three cases where lymphadenectomy was performed, the mean lymph node yield was 9, which is considered optimal for balancing the biochemical recurrence-free rate and complication risk.13 These findings indicate that the system can achieve oncological outcomes in radical prostatectomy comparable to those of established robotic systems.

Compared with our previously reported outcomes using other robotic systems,14 the performance of the current system was highly comparable. The mean operative time and hospital stay were 184.5 minutes/3.1 days (Sentire C1000), 225.8 minutes/3.3 days,14 (da Vinci S), and 223.7 minutes/3.0 days,14 (da Vinci SP). These results not only demonstrate that the Sentire C1000 offers performance comparable to existing robotic systems but also confirm the ease of surgical transition. A key factor enabling this transition is the similarity of the robotic control interface—including hand controls and foot pedals, allowing surgeons to adopt the new system more readily and apply their existing robotic experience. This is akin to driving different brands of cars, where consistent interfaces such as the steering wheel and pedals enable smooth adaptation. As a result, most new robotic systems adopt similar interface designs to support seamless integration by experienced surgeons. Moreover, the PSR has a similar arm configuration and setup to the da Vinci system, which also helps both surgeons and nursing staff adapt to the setup.

As this is the initial report of the Sentire Surgical System’s clinical use in radical prostatectomy, further clinical studies are warranted to evaluate its performance across a broader range of procedures.

Our findings indicate that the Sentire Surgical System (C1000) is a safe and effective robotic platform for radical prostatectomy. Its successful performance in this first-in-human experience supports ongoing development and broader application of this novel robotic surgical system.

Concept or design: CF Ng.
Acquisition of data: CH Yee, PKF Chiu, MHM Tam, FPT Lai.
Analysis or interpretation of data: FPF Lai.
Drafting of the manuscript: CF Ng, FPF Lai.
Critical revision of the manuscript for important intellectual content: All authors.

All authors had full access to the data, contributed to the study, approved the final version for publication, and take responsibility for its accuracy and integrity.

As an editor of the journal, CF Ng was not involved in the peer review process. Other authors have disclosed no conflicts of interest.

The authors thank the support from the operating theatre staff in Prince of Wales Hospital.

The research was supported by Cornerstone Robotics Limited for equipment and research fund. The funder had no role in the study design, data collection/analysis/interpretation, or manuscript preparation.

The research was approved by the Joint Chinese University of Hong Kong—New Territories East Cluster Clinical Research Ethics Committee, Hong Kong (Ref No.: CREC 2021.472). It was registered on www.ClinicalTrials.gov (Ref No.: NCT05151835). Written informed consent was obtained from all patients for the publication of this research and the accompanying images.

1. Ng AT, Tam PC. Current status of robot-assisted surgery. Hong Kong Med J 2014;20:241-50. Crossref

2. Okegawa T, Omura S, Samejima M, et al. Laparoscopic radical prostatectomy versus robot-assisted radical prostatectomy: comparison of oncological outcomes at a single center. Prostate Int 2020;8:16-21. Crossref

3. Allan C, Ilic D. Laparoscopic versus robotic-assisted radical prostatectomy for the treatment of localised prostate cancer: a systematic review. Urol Int 2016;96:373-8. Crossref

4. Bansal E, Kunaprayoon S, Zhang LP. Opportunities for global health diplomacy in transnational robotic telesurgery. AMA J Ethics 2023;25:E624-36. Crossref

5. Mehta A, Cheng Ng J, Andrew Awuah W, et al. Embracing robotic surgery in low- and middle-income countries: potential benefits, challenges, and scope in the future. Ann Med Surg (Lond) 2022;84:104803. Crossref

6. IDEAL Collaboration. The IDEAL Framework. Available from: https://www.ideal-collaboration.net/the-ideal-framework/. Accessed 26 May 2025.

7. Chan SY, Hou SM, Wong WS, Ng CF. Robotic urological surgery: prospects for Hong Kong. Surg Pract 2007;11:154-8. Crossref

8. Surgical Outcomes Monitoring and Improvement Programme (SOMIP) report. Volume Fifteen: July 2022—June 2023. Hospital Authority, Hong Kong SAR Government; 2024.

9. Lo KL, Ng CF, Lam CN, Hou SS, To KF, Yip SK. Short-term outcome of patients with robot-assisted versus open radical prostatectomy: for localised carcinoma of prostate. Hong Kong Med J 2010;16:31-5.

10. Marchegiani F, Siragusa L, Zadoroznyj A, et al. New robotic platforms in general surgery: what’s the current clinical scenario? Medicina (Kaunas) 2023;59:1264. Crossref

11. Novara G, Ficarra V, Rosen RC, et al. Systematic review and meta-analysis of perioperative outcomes and complications after robot-assisted radical prostatectomy. Eur Urol 2012;62:431-52. Crossref

12. Wu S, Lin SX, Cornejo KM, et al. Clinicopathological and oncological significance of persistent prostate-specific antigen after radical prostatectomy: a systematic review and meta-analysis. Asian J Urol 2023;10:317-28. Crossref

13. Zhang X, Zhang G, Wang J, Bi J. Different lymph node dissection ranges during radical prostatectomy for patients with prostate cancer: a systematic review and network meta-analysis. World J Surg Oncol 2023;21:80. Crossref

14. Ng CF, Teoh JY, Chiu PK, et al. Robot-assisted single-port radical prostatectomy: a phase 1 clinical study. Int J Urol 2019;26:878-83. Crossref

Declared a public health emergency of international concern by the World Health Organization, coronavirus disease 2019 (COVID-19) has continuously posed a threat to both physical and psychological health.1 Beginning in December 2021, the Omicron variant triggered the fifth wave of the pandemic in Hong Kong, endangering psychological well-being.1 2 Filipina domestic workers (FDWs), the primary group of migrant domestic workers, constitute >2.5% of the Hong Kong population3 and are considered a vulnerable population. Before the Hong Kong government reiterated the rights of migrant workers, many FDWs faced mistreatment, including abuse, exploitation, and illegal dismissal upon infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2).4 5 6 Filipina domestic workers were susceptible to both direct and indirect consequences of the COVID-19 pandemic.

Migrant workers often experience poor psychosocial conditions and substandard working environments.4 5 6 7 8 However, few studies have consistently examined the well-being of FDWs.8 9 10 Anxiety, a key indicator of well-being, commonly coexists with other psychological conditions. Considering the large number of domestic workers in Hong Kong, efforts to safeguard the psychological health of this minority population are essential to prevent excessive strain on the healthcare system.11 Additionally, various aspects of working conditions should be investigated in relation to anxiety.12

This study aimed to estimate the prevalence of anxiety and examine its relationships with perceived social support and decent work among FDWs after the peak of the Omicron wave during the COVID-19 pandemic in Hong Kong. Insights regarding the psychosocial conditions encountered by FDWs during the aftermath of the pandemic may contribute to existing literature.

A cross-sectional survey, written in English, was administered between June and August 2022. The target population comprised FDWs. Eligibility criteria included age ≥18 years, ability to read and understand English, and ability to provide informed consent. Filipina domestic workers who began employment on or after 1 February 2022 in Hong Kong, as well as male FDWs, were excluded from the present study. Because the majority of FDWs are women (97.8%), the inclusion of a small sample of male FDWs could compromise representativeness.3

Convenience sampling was utilised. Recruitment was conducted at gathering places in Central, Hong Kong, where a large proportion of FDWs spend their days off. Data collection was performed on rest days (Sundays and statutory holidays). Support and clarifications were provided to respondents who required assistance in understanding the questions. Respondents were offered a gratuity of HK$20 in cash as a token of appreciation for their time and assistance. According to Yeung et al,10 the prevalence of anxiety among FDWs in Hong Kong at the beginning of the pandemic was 25%. With a 95% confidence interval (95% CI) and a desired margin of error of ±5%, the minimum required sample size was estimated to be 289.

The questionnaire consisted of four sections, namely, anxiety, perceived social support, working conditions, and potential covariates (eg, socio-demographic and health-related factors). The questionnaire was developed based on validated instruments and a literature review of similar contexts.12 13 14 15 16 17

The Generalised Anxiety Disorder-7 (GAD-7) scale was adopted to assess anxiety levels.13 The total score ranges from 0 to 21; a threshold score of ≥10 to identify self-reported anxiety provides optimal sensitivity (89%) and specificity (82%).13 The GAD-7 has demonstrated high internal consistency in the general population (Cronbach’s alpha=0.92) and among FDWs working in Chinese regions (Cronbach’s alpha=0.80).18 19

The Multidimensional Scale of Perceived Social Support, using a 7-point Likert scale, was used to measure perceived social support across three domains, namely, significant others, family, and peers.14 Each domain comprises four items. We calculated a mean score for each domain ranging from 1 to 7 and a total mean score averaged from the three concerned domains to represent the total score of perceived social support. A higher score indicates a greater level of perceived social support. The authors of the scale proposed multiple approaches for interpreting perceived social support, one of which involved analysing continuous data for the three domains and the overall score.14 This scale has been validated with high internal consistency among Southeast Asian domestic workers in Hong Kong (Cronbach’s alpha=0.96).16 20

The Decent Work Scale was adopted to evaluate working conditions, including 15 items grouped into five components, namely, physically and interpersonally safe working conditions, access to essential healthcare support, sufficient income, adequate rest time, and alignment of working settings with social values.12 Each item scored from 1 to 7, resulting in component scores ranging from 3 to 21 and a total score ranging from 15 to 105, with higher scores indicating better working conditions. This scale has been validated with high internal consistency among the working population in the US (Cronbach’s alpha=0.86).12

Statistical analysis was performed using SPSS (Windows version 27.0; IBM Corp, Armonk [NY], US). Confidence intervals were established at the 95% level, and P values <0.05 were considered statistically significant. We computed 95% CIs for anxiety prevalence. Socio-demographic variables were compared between anxiety statuses using the Chi squared test, whereas scores for perceived social support and working conditions were compared using the independent samples t test.

Odds ratios (ORs) with 95% CIs were computed using a binary logistic regression model. For univariable analysis, simple logistic regression was conducted; perceived social support and working conditions constituted the main independent variables. Multivariable logistic regression analysis was performed to estimate the independent effects of these variables while adjusting for potential confounders.

The GAD-7 scores were categorised into four levels of anxiety severity: minimal (0-4), mild (5-9), moderate (10-14), and severe (15-21).13 We conducted sensitivity analysis using the GAD-7 score as an ordinal outcome and constructed an ordinal logistic regression model to assess the robustness of previously identified anxiety-associated factors.

Among the 441 FDWs approached, 71 declined to participate, yielding a response rate of 83.9% (Fig 1). Primary reasons for refusal were survey length and time constraints. The distribution of GAD-7 scores was positively skewed (Fig 2). The estimated prevalence of probable anxiety (GAD-7 score ≥10) was 8.6% (95% CI=5.8-11.5). Among the 370 respondents, approximately half were aged 35 to 44 years (51.1%) and married (48.4%). Most respondents had attained a university-level education or higher (60.3%), reported a monthly income ranging from HK$4630 to HK$4999 (68.6%), and had children residing in their home country (82.7%). The proportions of respondents residing on Hong Kong Island, in Kowloon and the New Territories were evenly distributed. The median (interquartile range) duration of employment in Hong Kong was 5.0 years (interquartile range, 3.0-9.0). Most respondents had no history of COVID-19 (81.9%) and no chronic diseases (97.8%) [Table 1]. Table 2 shows that the mean scores for the three domains of perceived social support ranged from 5.5 to 5.7 out of 7, whereas the mean score for decent work was 78.1 out of 105. Among the five components of working conditions measured by the Decent Work Scale, the lowest mean score was observed for rest periods (14.1); access to healthcare had the highest mean score (17.1).

Participants with probable anxiety had a higher proportion of chronic diseases relative to those without anxiety (9.4% vs 1.5%; P=0.024) [Table 1]. Respondents with probable anxiety reported worse perceptions of social support and working conditions; they had lower scores across all domains relative to those of respondents without anxiety (Table 2).

Simple logistic regression analysis indicated that one domain of perceived social support and multiple subscales of working conditions were significantly associated with anxiety (Table 3). Filipina domestic workers with higher perceived social support from significant others, better access to healthcare, greater satisfaction with compensation and salary, increased free time and rest periods, and higher satisfaction with their employer’s value orientation exhibited a lower likelihood of experiencing probable anxiety. Multivariable logistic regression analysis—adjusted for all relevant socio-demographic variables, health status, and subscales of perceived social support and working conditions—identified three variables that remained statistically significant (Table 3). Greater satisfaction with compensation and salary (adjusted odds ratio [aOR]=0.825, 95% CI=0.728-0.935), increased free time and rest periods (aOR=0.878, 95% CI=0.780-0.987), and higher satisfaction with value orientation (aOR=0.887, 95% CI=0.796-0.989) were associated with lower anxiety risk. Sensitivity analysis, which examined the four levels of anxiety as an ordinal outcome using an ordinal logistic regression model, showed that effect estimates were slightly attenuated. However, the findings confirmed the association between anxiety levels and inadequate compensation, while also identifying a history of chronic diseases as a risk factor for increased anxiety severity (Table 4).

The observed prevalence of anxiety among FDWs was 8.6%, representing a lower proportion compared with previous studies.10 11 21 22 The Omicron variant led to an unprecedented surge in cases, which peaked in early March 2022. Compared with a local study conducted at the onset of the COVID-19 pandemic,10 the prevalence of probable anxiety among FDWs declined from 25% to 8.6%. A remarkably lower prevalence of anxiety was observed when using the official cut-off score of ≥7 for the Anxiety subscale of the Depression, Anxiety, and Stress Scale-21 Items (DASS-21-A) in both the general population of Hong Kong (14%)11 and the Philippines (38.4%).21 In Singapore, 17.5% of migrant workers exhibited probable anxiety (DASS-21-A score ≥8).22 The discrepancy in anxiety prevalence across studies may be attributed to differences in study contexts and timeframes. Although the fifth wave of COVID-19 had nearly subsided in Hong Kong during the present study period, other regions were still experiencing high caseloads. The relatively low prevalence of anxiety among FDWs may indicate the development of psychological resilience after the Omicron pandemic. Additionally, information dissemination and vaccine availability were more established compared with the second and third waves of the pandemic.10

In response to the fifth wave of the COVID-19 pandemic, the local government implemented comprehensive public health policies to safeguard rights and facilitate risk communication among minority populations in Hong Kong. Coronavirus disease 2019 and vaccine-related information were made available in multiple languages, including Tagalog and English, thereby improving access to formal and accurate health information for FDWs. Access to adequate and accurate health information is essential for mitigating psychological distress and reducing anxiety levels associated with the pandemic, as demonstrated by the findings of a study conducted in the Philippines.21

Access to COVID-19 vaccines may partially explain the findings. In Hong Kong, domestic workers were designated as a priority group for vaccination within 1 month of launching the COVID-19 vaccination programme.23 Furthermore, the initial procurement of 22.5 million vaccine doses ensured sufficient supply for the entire population, allowing domestic workers to choose between Sinovac and BioNTech vaccines at no cost. The high effectiveness of COVID-19 vaccination may have contributed to anxiety reduction. As of August 2021, the majority of sampled domestic workers (80%) had received at least one dose of COVID-19 vaccine.24 A study by McMenamin et al25 demonstrated the substantial protective effect of COVID-19 vaccines against severe or fatal outcomes (BioNTech: two doses=83.9%; three doses=97.9%). Vaccination significantly reduces the risk of severe COVID-19 complications, hospitalisation, and mortality, which may have indirectly alleviated probable anxiety among FDWs. This assumption is supported by the results of a study examining the psychological impact of COVID-19 vaccination, which revealed lower anxiety levels among vaccinated individuals.26 However, the aforementioned local10 11 20 and Singapore studies22 assessing the anxiety of migrant workers were conducted during periods when no pharmaceutical preventive measures were available. Therefore, access to COVID-19 vaccines is a plausible explanation for the lower prevalence of probable anxiety among FDWs.

Additionally, job security may explain the decline in probable anxiety. Some FDWs expressed concerns regarding job insecurity and experienced distress due to job loss.4 Amid increasing reports of illegal contract terminations, the government intervened to uphold FDWs’ employment rights.27 On 5 March 2022, a government spokesperson emphasised zero tolerance for employers who illegally dismissed FDWs exhibiting SARS-CoV-2 infection.27 Any violation of the Employment Ordinance and related laws was subject to prosecution and fines.27 Filipina domestic workers exhibiting SARS-CoV-2 infection or identified as close contacts of individuals with COVID-19 receive the same assistance and support as other Hong Kong citizens, including quarantine and isolation arrangements.27 Greater institutional support for their employment may have contributed to the lower prevalence of anxiety among FDWs.

The significant others domain of perceived social support was negatively associated with anxiety in univariable analysis but was no longer significant according to multivariable regression. Significant others are individuals that the respondents regard as special persons.12 This finding contrasts with previous studies that identified perceived social support as an essential factor in coping with psychological distress among migrant workers.8 9 This discrepancy may be attributable to the small sample size. However, the finding is consistent with results from a local study conducted in a similar context.10

Filipina domestic workers migrate to foreign countries to support their families’ livelihoods; they are often portrayed as resilient and independent figures by the Philippine Government. This narrative may subtly reinforce the perception among FDWs that they are the sole breadwinners responsible for their families’ well-being.28 Consequently, although FDWs may seek informal social support from significant others, their self-disclosure remains selective. Psychological concerns, in particular, may be considered sensitive topics, leading to avoidance of such discussions in an effort to protect their self-esteem. This avoidance may explain the absence of an observed association between perceived social support and anxiety.

Another key finding was that better working conditions—including greater satisfaction with compensation and salary, increased free time and rest periods, and higher satisfaction with value orientation—were associated with a lower likelihood of probable anxiety. Working conditions are recognised as social determinants of mental health. Findings from the World Health Organization suggest that jobs offering high rewards and a greater sense of control serve as protective factors for mental well-being, thereby reinforcing the importance of favourable working conditions for employees.29 Consistent with the previous findings,30 high and regular monetary compensation was linked to lower probable anxiety in our study. According to the Occupational Wages Survey in the Philippines,30 the median monthly income was PHP13 646 (HK$1865, US$239), whereas the minimum monthly wage in Hong Kong was HK$4630 (US$594) during the study period.31 Filipina domestic workers in Hong Kong earned at least 2.48-fold more than their counterparts in the Philippines. Higher monthly earnings are often allocated toward property purchases in the Philippines, meeting family obligations, and fulfilling roles and responsibilities. Thus, greater satisfaction with compensation and salary may have contributed to lower probable anxiety among FDWs. Although this factor may explain the observed association, a qualitative study would provide deeper insights into the relationship between higher compensation and reduced psychological distress.

Additionally, increased free time and rest periods were associated with a lower risk of probable anxiety. An occupational health study32 established an inverse relationship between working hours and sleep duration, where anxiety and depression scores were higher among individuals working longer hours. These findings suggest that increased free time and rest periods can help reduce anxiety risk.

Notably, greater alignment between FDWs’ working environments and their social values was associated with lower anxiety risk. Value orientation refers to the principles an individual upholds, including ethics, morality, and attitudes toward work. In the workplace, each aspect of the working environment is interconnected with FDWs and their employers, influencing the likelihood of psychological distress. Employers are encouraged to engage in discussions with FDWs regarding working conditions—such as job demands and task restructuring—to ensure alignment in value orientation between both parties.

While chronic disease was not a statistically significant predictor of anxiety in multivariable logistic regression model, sensitivity analysis using an ordinal outcome revealed that it remained a risk factor for increased anxiety severity. Despite the inconclusive findings regarding this association, a systematic review33 indicated that a history of chronic diseases is linked to higher anxiety levels. The presence of chronic diseases has a negative impact on mental health.33

Some limitations were inherent in our sampling method and study design. First, we could not establish causality. Because cross-sectional study designs provide only short-term data regarding associations, longitudinal studies are needed to examine temporal sequences and causal relationships. Second, the use of convenience sampling may introduce selection bias; therefore, generalisations of the findings to the entire FDW population should be made with caution. However, this bias is likely minimal because all FDWs were approached, and none were selectively invited based on specific characteristics; also, the demographic distribution of the sample closely resembled that of domestic workers recorded in the Hong Kong Population Census.34 The age distributions in the Census data34 and the study sample were comparable: 18-34 years (29.8% vs 27.8%), 35-44 years (48.2% vs 51.1%), and ≥45 years (22.0% vs 21.1%). Additionally, the respondents’ residence areas were evenly distributed across Hong Kong Island, Kowloon, and the New Territories. These findings suggest high representativeness and generalisability in the study sample. Furthermore, monetary incentives were provided, which may have contributed to higher-quality responses.

This study identified associations between optimal working conditions and lower probable anxiety among FDWs. The findings update the estimated prevalence of anxiety in this population and suggest that favourable working conditions may serve as protective factors. The study provides insights for the development and refinement of public health measures and occupational policies related to migrant workers, including compensation for overtime work, job security, and adequate rest periods. Psychological interventions tailored to domestic workers should be developed to address diverse mental health needs while incorporating labour protection. Regular review and refinement of occupational policies may be necessary. The Labour Department could consider conducting large-scale quantitative surveys and qualitative interviews with domestic workers to assess and accommodate their occupational needs. Future studies should aim to include domestic workers of various nationalities and other migrant worker populations.

Concept or design: KKY Lai, ELY Wong.
Acquisition of data: KKY Lai.
Analysis or interpretation of data: All authors.
Drafting of the manuscript: KKY Lai.
Critical revision of the manuscript for important intellectual content: ELY Wong.

All authors had full access to the data, contributed to the study, approved the final version for publication, and take responsibility for its accuracy and integrity.

All authors have disclosed no conflicts of interest.

The authors thank Prof Marc KC Chong, Ms Annie WL Cheung and Mr Jonathan CH Ma from the Centre for Health Systems and Policy Research, The Jockey Club School of Public Health and Primary Care, The Chinese University of Hong Kong for their valuable comments on the study and support in data analysis. The authors also thank all study respondents for their valuable time in completing the questionnaires and for their contributions as migrant workers in Hong Kong.

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

This research was approved by the Survey and Behavioural Research Ethics Committee of The Chinese University of Hong Kong, Hong Kong (Ref No.: 018-22). The study was conducted in accordance with the principles of the Declaration of Helsinki. Informed consent was obtained from the participants prior to commencement of the survey.

1. Centre for Health Protection and Hospital Authority, Hong Kong SAR Government. Statistics on 5th wave of COVID- 19 (from 31 Dec 2021 up till 31 May 2022 00:00). Available from: https://www.coronavirus.gov.hk/pdf/5th_wave_statistics/5th_wave_statistics_20220531.pdf. Accessed 5 Dec 2022.

2. Xiong J, Lipsitz O, Nasri F, et al. Impact of COVID-19 pandemic on mental health in the general population: a systematic review. J Affect Disord 2020;277:55-64. Crossref

3. Census and Statistics Department, Hong Kong SAR Government. 2021 Population Census. Main Results. Available from: https://www.census2021.gov.hk/doc/pub/21c-main-results.pdf. Accessed 1 Apr 2025.

4. Chow Y. No home away from home for domestic workers terminated after contracting coronavirus amid Hong Kong’s fifth wave. Young Post. South China Morning Post; 2022 May 16. Available from: https://www.scmp.com/yp/discover/news/hong-kong/article/3177657/no-home-away-home-domestic-workers-terminated-after. Accessed 4 Dec 2022.

5. Cheung JT, Tsoi VW, Wong KH, Chung RY. Abuse and depression among Filipino foreign domestic helpers. A cross-sectional survey in Hong Kong. Public Health 2019;166:121-7. Crossref

6. Choy CY, Chang L, Man PY. Social support and coping among female foreign domestic helpers experiencing abuse and exploitation in Hong Kong. Front Commun 2022;7:1015193. Crossref

7. Sterud T, Tynes T, Mehlum IS, et al. A systematic review of working conditions and occupational health among immigrants in Europe and Canada. BMC Public Health 2018;18:770. Crossref

8. Ioannou M, Kassianos AP, Symeou M. Coping with depressive symptoms in young adults: perceived social support protects against depressive symptoms only under moderate levels of stress. Front Psychol 2019;9:2780. Crossref

9. Straiton ML, Aambø AK, Johansen R. Perceived discrimination, health and mental health among immigrants in Norway: the role of moderating factors. BMC Public Health 2019;19:325. Crossref

10. Yeung NC, Huang B, Lau CY, Lau JT. Feeling anxious amid the COVID-19 pandemic: psychosocial correlates of anxiety symptoms among Filipina domestic helpers in Hong Kong. Int J Environ Res Public Health 2020;17:8102. Crossref

11. Choi EP, Hui BP, Wan EY. Depression and anxiety in Hong Kong during COVID-19. Int J Environ Res Public Health 2020;17:3740. Crossref

12. Duffy RD, Allan BA, England JW, et al. The development and initial validation of the Decent Work Scale. J Couns Psychol 2017;64:206-21. Crossref

13. Spitzer RL, Kroenke K, Williams JB, Löwe B. A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch Intern Med 2006;166:1092-7. Crossref

14. Zimet GD, Dahlem NW, Zimet SG, Farley GK. The Multidimensional Scale of Perceived Social Support. J Pers Assess 1988;52:30-41. Crossref

15. International Organization for Migration. The Determinants of Migrant Vulnerability. Geneva: United Nations; 2019. Available from: https://www.iom.int/sites/g/files/tmzbdl486/files/our_work/DMM/MPA/1-part1-thedomv.pdf. Accessed 5 Dec 2022.

16. Garabiles MR, Lao CK, Yip P, Chan EW, Mordeno I, Hall BJ. Psychometric validation of PHQ-9 and GAD-7 in Filipino migrant domestic workers in Macao (SAR), China. J Pers Assess 2020;102:833-44. Crossref

17. Mendoza NB, Mordeno IG, Latkin CA, Hall BJ. Evidence of the paradoxical effect of social network support: a study among Filipino domestic workers in China. Psychiatry Res 2017;255:263-71. Crossref

18. Yeung NC, Kan KK, Wong AL, Lau JT. Self-stigma, resilience, perceived quality of social relationships, and psychological distress among Filipina domestic helpers in Hong Kong: a mediation model. Stigma Health 2021;6:90-9. Crossref

19. Pan American Health Organization. Questionnaire to Assess the Diagnosis and Treatment of Chronic Diseases. Geneva: World Health Organization. Available from: https://www.paho.org/hq/dmdocuments/2009/cncd_mgt_questionnaire.pdf. Accessed 4 Dec 2022.

20. Leung DD, Tang EY. Correlates of life satisfaction among Southeast Asian foreign domestic workers in Hong Kong: an exploratory study. Asian Pac Migr J 2018;27:368-77. Crossref

21. Tee ML, Tee CA, Anlacan JP, et al. Psychological impact of COVID-19 pandemic in the Philippines. J Affect Disord 2020;277:379-91. Crossref

22. Saw YE, Tan EY, Buvanaswari P, Doshi K, Liu JC. Mental health of international migrant workers amidst large-scale dormitory outbreaks of COVID-19: a population survey in Singapore. J Migr Health 2021;4:100062. Crossref

23. Labour Department, Hong Kong SAR Government. Foreign domestic helpers. Vaccination priority groups to be expanded to cover people aged 30 or above. 2021 Mar 15. Available from: https://www.fdh.labour.gov.hk/en/news_detail.html?year=2021&n_id=190. Accessed 28 Mar 2025.

24. Sumerlin TS, Kim JH, Wang Z, Hui AY, Chung RY. Determinants of COVID-19 vaccine uptake among female foreign domestic workers in Hong Kong: a cross-sectional quantitative survey. Int J Environ Res Public Health 2022;19:5945. Crossref

25. McMenamin ME, Nealon J, Lin Y, et al. Vaccine effectiveness of one, two, and three doses of BNT162B2 and CoronaVac against COVID-19 in Hong Kong: a population-based observational study. Lancet Infect Dis 2022;22:1435-43. Crossref

26. Babicki M, Malchrzak W, Hans-Wytrychowska A, Mastalerz-Migas A. Impact of vaccination on the sense of security, the anxiety of COVID-19 and quality of life among polish. A nationwide online survey in Poland. Vaccines (Basel) 2021;9:1444. Crossref

27. Hong Kong SAR Government. Government’s response on situation of foreign domestic helpers affected by COVID-19 (with photos) [press release]. 2022 Mar 5. Available from: https://www.info.gov.hk/gia/general/202203/05/P2022030500399.htm. Accessed 4 Dec 2022.

28. Rich GJ. Filipina migrant domestic workers in Asia: mental health and resilience. In: Rich GJ, Jaafar JL, Barron D, editors. Psychology in Southeast Asia: Sociocultural, Clinical, and Health Perspectives. London: Routledge, Taylor & Francis Group; 2020. Crossref

29. World Health Organization. Social determinants of mental health. 2014 May 18. Available from: https://www.who.int/publications/i/item/9789241506809. Accessed 1 Apr 2025.

30. Mapa DS. Average monthly wage rates of selected occupations: 2018 and 2020 [Internet]. 2020 Occupational Wages Survey (OWS). Philippine Statistics Authority; 2022. Available from: https://psa.gov.ph/statistics/occupational-wages-survey/node/168472. Accessed 4 Dec 2022.

31. Hong Kong SAR Government. Minimum allowable wage and food allowance for foreign domestic helpers [press release]. 2021 Sep 30. Available from: https://www.info.gov.hk/gia/general/202109/30/P2021093000329.htm. Accessed 4 Dec 2022.

32. Afonso P, Fonseca M, Pires JF. Impact of working hours on sleep and mental health. Occup Med (Lond) 2017;67:377-82. Crossref

33. Clarke DM, Currie KC. Depression, anxiety and their relationship with chronic diseases: a review of the epidemiology, risk and treatment evidence. Med J Aust 2009;190:S54-60. Crossref

34. Census and Statistics Department, Hong Kong SAR Government. 2021 Population Census: Summary Results. 2022. Available from: https://www.censtatd.gov.hk/en/data/stat_report/product/B1120106/att/B11201062021XXXXB01.pdf. Accessed 4 Dec 2022.