Many individuals are diagnosed with cancer during childhood, adolescence, and young adulthood. Worldwide, there were approximately 1 335 100 new cancer cases among adolescents and young adults in 20191; the incidence rate was 44.99 per 100 000 people.1 In 2020, the incidence rate for cancer among Hong Kong children and adolescents (aged 0-19 years) was 160 cases per 1 000 000 people.2 There were 177 newly diagnosed cancer cases in this age-group (92 in male patients and 85 in female patients).2 The survival rates for childhood and adolescent cancers are encouraging. In a retrospective cohort study from a research hospital in the United States,3 the 5-year overall survival rate exceeded 83%. Similarly, in Hong Kong, the 5-year survival rate among women diagnosed with breast cancer, the most common cancer in reproductive-age women, was 84% between 2010 and 2017.2

Chemotherapy or pelvic radiotherapy may affect fertility, either temporarily or permanently. Considering advances in cancer treatment and improved post-treatment survival rates, fertility should be discussed at the time of cancer diagnosis, especially for younger patients who have not yet completed their families. International guidelines regarding fertility preservation (FP) recommend that clinicians inform cancer patients about the potential effects of cancer and its treatment on reproductive function, as well as FP options.4 5 In a semi-structured phone interview study of female cancer survivors who were diagnosed with invasive cervical cancer, breast cancer, Hodgkin lymphoma, or non-Hodgkin lymphoma at age ≤40 years, participants were interviewed an average of 10 years after diagnosis.6 Those who had wanted children at the time of diagnosis but were unable to conceive subsequently reported distress related to their interrupted fertility.6 Additionally, patients who do not receive accurate and timely information regarding FP are at risk for psychological distress.7 In our recently published cross-sectional questionnaire study of reproductive-age women in Hong Kong who had been diagnosed with breast cancer,8 only 44% of those women were aware of FP; however, 46% of the women felt that fertility concerns affected their cancer treatment decisions.8

The most common FP options include sperm cryopreservation for men and embryo or oocyte cryopreservation for women. Other options for women include pharmacological ovarian protection using gonadotropin-releasing hormone (GnRH) agonists, ovarian tissue cryopreservation, and ovarian transposition. In Hong Kong, FP was previously self-funded and only available through private services. Sperm cryopreservation costs approximately HK$4400 to HK$6600 for 2 years, whereas oocyte and embryo cryopreservation costs are approximately HK$15 000 to HK$20 000.9 Our centre launched the first public FP programme for cancer patients in Hong Kong, beginning in August 2020. Here, we review the two-and-a-half-year experience of providing public FP services to cancer patients in Hong Kong.

This retrospective study included men and women who attended the Centre of Assisted Reproduction and Embryology at The University of Hong Kong—Queen Mary Hospital for FP services before cancer treatment, from the establishment of our public FP programme in August 2020 until the end of February 2023.

During the study period, we provided public FP for cancer patients <35 years old, expressed a desire for future fertility, had a survival rate exceeding 50% after cancer treatment, had no living children, and had not undergone prior chemotherapy or pelvic radiotherapy. In women, an antral follicle count of >7 on pelvic ultrasound was required. These criteria were adapted from The Edinburgh Selection Criteria for ovarian tissue cryopreservation.10

There was no minimum age requirement for FP. Male adolescents could undergo sperm freezing if they were able to provide sperm samples for cryopreservation. For patients aged <18 years, we included their parents in discussions prior to proceeding with FP treatment.

During the study period, the public FP programme offered up to 40 cycles of sperm freezing and 20 cycles of oocyte/embryo freezing per year.

Patients diagnosed with cancer who were expected to undergo gonadotoxic treatments were referred to our FP service by surgeons, oncologists, paediatricians, haematologists, private practitioners, and cancer support groups. Clinicians completed a referral letter, which can be downloaded from our centre’s website.11 Patients or their doctors can also contact us via email. Additionally, a chat group was established between Hong Kong Children’s Hospital and our centre to facilitate rapid referrals.

After we received a referral, the patient was scheduled for an appointment in the public FP clinic within 1 week. Our centre maintained a flexible clinic schedule, which allowed urgent cases to be accommodated within the existing clinic framework, 5 days per week.

The details of our FP programme were previously published.12 Information sheets and videos about the FP services offered by our centre were readily accessible to the general population and patients through our website13 and YouTube channel.14 Patients were encouraged to review these materials before attending the public FP clinic. For men, sperm banking was arranged on the same day as counselling. For women, the options of oocyte and embryo preservation were discussed if feasible. Embryo preservation was only offered to women who were legally married. In Hong Kong, assisted reproductive technology is regulated by the Human Reproductive Technology Ordinance.15 This ordinance limits the storage duration for frozen gametes in cancer patients to 10 years or until the patient reaches the age of 55 years, whichever is longer.15 The storage duration for frozen embryos is limited to 10 years.15 Cryopreserved gametes and embryos can only be used after a patient recovers from their illness and is legally married.15 Posthumous use of cryopreserved gametes and embryos is prohibited.15

The use of GnRH agonists for pharmacological ovarian protection was discussed either after cryopreservation or if cryopreservation was not feasible. Such agonists were usually administered monthly or every 3 months during chemotherapy.

The characteristics of men who underwent sperm cryopreservation and women who underwent ovarian stimulation for oocyte or embryo cryopreservation were prospectively entered into our database. Medical records (both in paper and electronic formats), including data from the assisted reproductive technology database at our centre and the Hospital Authority’s electronic clinical management system, were retrieved and reviewed. These records encompassed demographic data, cancer type, cancer treatment, FP method chosen, ovarian stimulation cycle characteristics, semen analysis, reproductive outcomes, and follow-up information, if available.

All women who attended our centre for FP were asked to return to our late-effects clinic for gonadal function monitoring after the completion of cancer treatment. All men were asked to undergo semen analysis when they wished to conceive after the completion of cancer treatment.

Data were analysed using SPSS (Windows version 26; IBM Corp, Armonk [NY], United States) and are presented as median (interquartile range [IQR]) or as number (percentage). P value was calculated by Chi squared test.

Fifty-two women were referred to our public FP clinic between August 2020 and February 2023. Three women were excluded from the analysis because they had non-malignant conditions, including rheumatological disease (systemic lupus erythematosus) and neurological disease (multiple sclerosis). Additionally, one woman missed her clinic appointment. Therefore, the final analysis included 48 women (Fig a). The median age of these women was 30 years (IQR=25-33). The cancer outcomes of these women are shown in Table 1. Regarding marital status, 36 women (75.0%) were single, 11 (22.9%) were married, and one (2.1%) was divorced. All were nulliparous, except for one married woman (2.1%) with a livebirth was ineligible for publicly funded FP due to the programme’s criteria. She then selected GnRH agonist treatment after counselling. The median time from referral to consultation was 3 days (IQR=2-5).

Eighteen women underwent 19 cycles of ovarian stimulation for oocyte or embryo cryopreservation (Table 2). One woman underwent an additional self-financed stimulation cycle because she only achieved two frozen oocytes in the first cycle. She achieved two additional frozen oocytes in the second attempt. Thirteen women cryopreserved oocytes, whereas five women cryopreserved embryos (three at the cleavage stage and two at the blastocyst stage). The median time between consultation and ovarian stimulation was 5 days (IQR=2-12) [Table 2]. All women with breast cancer received letrozole co-treatment during ovarian stimulation.

One woman developed moderate ovarian hyperstimulation syndrome requiring hospital admission. Oocyte retrieval was uneventful, and 45 oocytes were retrieved. However, 3 days after oocyte retrieval, she was admitted with abdominal distension, shortness of breath, and vomiting. She was diagnosed with moderate ovarian hyperstimulation syndrome, which resolved with conservative management.

There was no significant age difference between women who proceeded with oocyte/embryo cryopreservation and those who did not. The median age of women who proceeded with oocyte/embryo cryopreservation was 28 years (IQR=24.0-32.8), whereas the median age of women who did not proceed with oocyte/embryo cryopreservation was 31 years (IQR=26.8-33.0) [Table 3].

Among patients with breast and gynaecological cancers, six of 10 (60.0%) and six of 13 (46.2%) underwent oocyte/embryo cryopreservation, respectively, compared with five of 19 (26.3%) women with haematological cancers and one of six (16.7%) women with other solid tumours (Table 3).

Among the 48 women who attended the clinic, nine (18.8%) proceeded with oocyte or embryo cryopreservation alone, nine (18.8%) underwent cryopreservation followed by the use of GnRH agonists, 21 (43.8%) received GnRH agonists alone, five (10.4%) decided against FP after counselling, and four (8.3%) were lost to follow-up. Those who chose GnRH agonists received this treatment from their primary oncology team.

At the end of February 2023, among the 48 women, 22 exhibited disease remission, 21 were continuing treatment, four were deceased, and one had been lost to follow-up (Table 1). None of the women have returned to use their frozen oocytes or embryos, nor have any reported natural conception since their cancer diagnosis.

Sixty-six men were referred to our public FP clinic during the study period (Fig b). Five men were excluded: four had exceeded the age limit and one had already begun chemotherapy. Fertility preservation counselling at a private clinic was offered to those who were not eligible for the public FP service. One man, who exceeded the age limit, underwent self-funded sperm cryopreservation. Three men missed their clinic appointments. Therefore, the final analysis included 58 men (Fig b). The median age of the men was 26 years (IQR=18.3-32.8). The cancer outcomes of these men are shown in Table 4. Regarding marital status, 51 men (87.9%) were single and seven men (12.1%) were married. One man (1.7%) had a child but was unmarried. The remaining 57 men (98.3%) had no offspring. The median time from referral to consultation was 4 days (IQR=2-7).

Among the 58 men who attended the clinic, 55 attempted sperm freezing and three chose not to undergo cryopreservation after counselling. Six men were unable to cryopreserve sperm (Fig b). One, aged 14 years, was unable to provide a semen sample; four men submitted semen samples containing no sperm. One man had previously attempted sperm cryopreservation at a private hospital, but no sperm were found in his ejaculate. He subsequently underwent testicular sperm extraction at our hospital; no sperm were retrieved. The ages of the men with no sperm in their semen ranged from 15 to 34 years.

The median number of vials of cryopreserved sperm was 5 (IQR=5-5) and the median sperm concentration was 18.8 million/mL (IQR=4.3-52.8).

At the end of February 2023, among the 58 men, 29 exhibited disease remission, 18 were continuing treatment, six were deceased, and five had been lost to follow-up (Table 4). None of the men have returned to use their frozen sperm.

As of this writing, six men and four woman who attended the FP clinic have died.

This is the first review of a public FP programme for cancer patients in Hong Kong. Our study showed that among the 48 women who attended during the study period, 37.5% (n=18) underwent oocyte/embryo cryopreservation and 62.5% (n=30) chose GnRH agonists for FP. In contrast, among the 58 men who attended for FP before cancer treatment, >90% attempted sperm cryopreservation.

We previously published a review of our self-funded FP service from 2010 to 2020.12 During that period, 72 women attended consultations for FP, and 20 of them underwent 22 cycles of ovarian stimulation for oocyte or embryo cryopreservation.12 Additionally, from 1995 to 2020, 265 men underwent sperm cryopreservation.12 Over the years, there were increases in the numbers of men and women seeking FP; the increase was more prominent among women.

For comparison, we selected the period from 2018 to 2020 (ie, the 2.5 years immediately preceding the launch of the public FP programme). During that period, 19 women were referred for self-funded FP prior to cancer treatment, and 10 (52.6%) underwent oocyte or embryo cryopreservation. Fifty-eight men were referred for FP and underwent sperm cryopreservation. In the years prior to the launch of the publicly funded FP programme, we had already begun networking with various specialties, which likely contributed to the gradual increase in awareness and demand for FP services.

A successful FP programme requires good networking, flexibility, and a patient-friendly clinic environment. During the establishment of the public FP programme, we have networked with other specialties to enhance collaboration. Our centre aimed to simplify logistics so that consultations could be arranged as quickly as possible, allowing FP counselling and procedures to be completed within the short window of opportunity before cancer treatment. In our public FP clinic, the median waiting times from referral to consultation were 3 days for women and 4 days for men. Among women who chose oocyte or embryo cryopreservation, the median time from consultation to the start of ovarian stimulation was 5 days (IQR=2-12). In our previous study, the time from consultation to oocyte retrieval was 17 days (IQR=13-30).12 Notably, our previous study did not investigate the waiting time from referral to consultation; therefore, direct comparisons cannot be performed. Compared with our previous study regarding FP for cancer patients at our centre,12 the proportion of women who ultimately underwent oocyte or embryo cryopreservation increased from 28% to 38% in the public FP programme. However, further monitoring is needed to determine whether this difference represents a true upward trend due to increased awareness and easier access to the service. Additionally, patient characteristics and cancer types may vary across time periods.

For reproductive-age women with cancer, the receipt of specialised counselling regarding fertility issues, followed by FP, has been linked to less regret and improved quality of life among survivors.16 Providing our patients with accessible FP counselling and affordable treatments is an essential aspect of comprehensive oncology care. A clinical practice guideline from the American Society of Clinical Oncology indicates that FP should be initiated as early as possible in the treatment process to allow for the widest range of options.17 Referral to FP services enables patients to receive counselling from reproductive medicine specialists, empowering them to make informed decisions about fertility treatment.

At our FP clinic, patients were able to consult reproductive medicine specialists who discussed the potential effects of gonadotoxic cancer treatments on future fertility and described FP options. Local regulations concerning gamete storage and assisted reproduction were also explained. Patients were informed that they must be legally married to use frozen gametes in the future, and that gametes cannot be used posthumously.

In our cohort, only 62.5% of women received GnRH agonists during cancer treatment; the reasons for not receiving GnRH agonists were not recorded. Gonadotropin-releasing hormone agonists are usually administered monthly or every 3 months during cancer treatment, although their effectiveness depends on the type of cancer treatment. Some studies of breast cancer patients have shown that GnRH agonists can reduce the risk of premature ovarian insufficiency, but the fertility benefit remains uncertain.18 19 20 Most studies have focused on outcomes such as the maintenance or resumption of menstruation, prevention of treatment-related premature ovarian failure, and ovulation. In a Cochrane review20 which discussed randomised controlled trials that examined the effect of GnRH analogues for chemotherapy-induced ovarian failure in premenopausal women, 12 randomised controlled trials were included. Eleven studies reported rates of menstruation recovery or maintenance, four studies measured treatment-related premature ovarian failure, and seven studies reported the rates of pregnancy.20 However, there are limited data regarding live birth rates.20 A meta-analysis of randomised studies concerning ovarian suppression using GnRH agonists during chemotherapy in breast cancer patients found that temporary ovarian suppression with a GnRH agonist in young breast cancer patients was associated with a reduced risk of chemotherapy-induced premature ovarian insufficiency; it also appeared to increase the pregnancy rate without negatively influencing prognosis.21 Thus far, the benefit of GnRH agonists in other malignancies is unclear. A long-term analysis of young female lymphoma patients showed that GnRH agonists were not effective in preventing chemotherapy-induced premature ovarian insufficiency and did not improve future pregnancy rates.22 According to the European Society of Human Reproduction and Embryology guideline on female FP,4 GnRH agonists should be offered as an option for protecting ovarian function in premenopausal breast cancer patients undergoing chemotherapy; importantly, limited evidence exists regarding their protective effects on ovarian reserve and potential future pregnancies.4 In malignancies other than breast cancer, GnRH agonists should not be routinely offered as an option for protecting ovarian function protection and FP without discussing the uncertainty of their benefit.4 Gonadotropin-releasing hormone agonists during chemotherapy should not be considered as a substitute for established FP techniques, such as cryopreservation. They can be offered in addition to cryopreservation or when such techniques are not feasible.4 Despite the use of GnRH agonists, patients may experience premature ovarian insufficiency. Gonadotropin-releasing hormone agonists are currently provided as a self-financed option; women are often referred back to their oncology team, who prescribes and administers these agonists after FP counselling. The proportion of patients who underwent oocyte/embryo cryopreservation was higher among those with gynaecological or breast cancers than among those with haematological malignancies. This difference is likely due to the nature of their diseases and the urgency of initiating cancer treatment.

For women who chose to proceed with ovarian stimulation for oocyte or embryo cryopreservation, oocytes were retrieved during a stimulated cycle. Recombinant follicle-stimulating hormone could be initiated on any day of the menstrual cycle for ovarian stimulation (ie, ‘random-start’), using either a GnRH antagonist or progestin-primed protocol. This random-start approach allowed ovarian stimulation without substantial delays and did not affect the number or quality of retrieved oocytes.4 For women with hormone-sensitive cancers (eg, breast cancer), letrozole was routinely used during ovarian stimulation. The concomitant use of letrozole reduced circulating oestrogen levels and did not impair the efficacy of ovarian stimulation.23 A systematic review and meta-analysis regarding the safety of hormonal stimulation in young women with breast cancer before starting cancer treatment, as well as survivors who underwent assisted reproduction after cancer treatment, showed no increased risk of breast cancer recurrence in women who underwent ovarian stimulation with concomitant letrozole treatment.24 Despite using the ‘random-start’ approach, one cycle of ovarian stimulation required approximately 2 weeks.

This study had some limitations. It was a retrospective, single-centre study conducted over a short period of time; thus, it may not reflect situations in other regions. Due to resource constraints, we only included cancer patients who had not begun cancer treatment. Patients who did not meet the criteria for the public FP programme but still wished to pursue FP were referred to private clinics or other private centres upon receipt of their referral and therefore were excluded from this review. Patients who had already begun cancer treatment were also excluded from the public FP service. However, they could still be referred to our centre after stabilisation to assess fertility and explore self-funded FP options before undergoing more toxic chemotherapy, non–fertility-sparing radiotherapy, or surgeries. At the time of writing, our centre has not yet offered ovarian or testicular tissue cryopreservation. A 2018 survey of several Asian countries (eg, Australia, China, and India) revealed that ovarian tissue cryopreservation was available for prepubertal girls and postpubertal women who were unable to delay the initiation of chemotherapy.25 Testicular tissue cryopreservation also was provided to prepubertal boys in Australia, China, India, Indonesia, Japan, and Taiwan.25 A recently published pilot study from Hong Kong demonstrated the feasibility of ovarian tissue cryopreservation and transplantation using xenografts in nude mice26; ovarian tissue cryopreservation has recently become available in Hong Kong.

The patients included in this study were counselled for FP, and many are still undergoing cancer treatment and monitoring; none have returned to use the stored material. They were advised to return after cancer treatment for follow-up regarding their gonadal function. Patient satisfaction should also be evaluated. However, at the time of cryopreservation—typically close to the time of cancer diagnosis—patients may feel overwhelmed by the diagnosis and planned cancer treatments. Thus, patient satisfaction may be more accurately evaluated when the cancer is controlled or in remission.

Despite the presence of the public FP programme, patients were required to pay for the medications used in ovarian stimulation, as well as the fees involved in oocyte handling, freezing, and storage of frozen gametes or embryos; these costs were considerably reduced compared with expenses in private clinics. Cost remains a major barrier to accessing FP services. Although a public healthcare system has been established in Hong Kong, cancer patients are often financially overwhelmed due to the loss of income after a cancer diagnosis, along with additional expenditures for various self-funded investigations or treatments. We recently performed a survey of the knowledge, attitudes, and intentions regarding FP among breast cancer patients; most participants thought that FP should be subsidised by the government or provided at no cost.8

Since the establishment of a public FP programme for cancer patients, there has been an increase in the number of patients seeking FP services. More than 90% of men attempted sperm cryopreservation, whereas 37.5% of women underwent oocyte/embryo cryopreservation and 62.5% of women received GnRH agonists during cancer treatment. With further promotion, changes in funding policies, and a more accessible FP programme, the demand for FP services is expected to increase. Fertility preservation services should be regularly reviewed to assess changes in demand and identify areas for improvement.

Concept or design: JKY Ko, EHY Ng.
Acquisition of data: DTY Chan.
Analysis or interpretation of data: DTY Chan, JKY Ko, EHY Ng.
Drafting of the manuscript: DTY Chan.
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.

All authors have disclosed no conflicts of interest.

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

The study protocol was approved by the Institutional Review Board of The University of Hong Kong/Hospital Authority Hong Kong West Cluster, Hong Kong (Ref No.: UW 23-334). The requirement for informed consent was waived by the Board due to the retrospective nature of the research.

1. You L, Lv Z, Li C, et al. Worldwide cancer statistics of adolescents and young adults in 2019: a systematic analysis of the Global Burden of Disease Study 2019. ESMO Open 2021;6:100255. Crossref

2. Hong Kong Cancer Registry, Hospital Authority. Overview of Hong Kong Cancer Statistics of 2020. Available from: https://www3.ha.org.hk/cancereg/pdf/overview/Overview%20of%20HK%20Cancer%20Stat%202020.pdf. Accessed 7 May 2023.

3. Bhakta N, Liu Q, Ness KK, et al. The cumulative burden of surviving childhood cancer: an initial report from the St Jude Lifetime Cohort Study (SJLIFE). Lancet 2017;390:2569-82. Crossref

4. ESHRE Guideline Group on Female Fertility Preservation; Anderson RA, Amant F, et al. ESHRE guideline: female fertility preservation. Hum Reprod Open 2020;2020:hoaa052. Crossref

5. Practice Committee of the American Society for Reproductive Medicine. Fertility preservation in patients undergoing gonadotoxic therapy or gonadectomy: a committee opinion. Fertil Steril 2019;112:1022-33. Crossref

6. Canada AL, Schover LR. The psychosocial impact of interrupted childbearing in long-term female cancer survivors. Psychooncology 2012;21:134-43. Crossref

7. Rosen A, Rodriguez-Wallberg KA, Rosenzweig L. Psychosocial distress in young cancer survivors. Semin Oncol Nurs 2009;25:268-77. Crossref

8. Ko JK, Cheung CS, Cheng HH, et al. Knowledge, attitudes and intention on fertility preservation among breast cancer patients. Sci Rep 2023;13:9645. Crossref

9. Yeung SY, Ng EY, Lao TT, Li TC, Chung JP. Fertility preservation in Hong Kong Chinese society: awareness, knowledge and acceptance. BMC Womens Health 2020;20:86. Crossref

10. Wallace WH, Smith AG, Kelsey TW, Edgar AE, Anderson RA. Fertility preservation for girls and young women with cancer: population-based validation of criteria for ovarian tissue cryopreservation. Lancet Oncol 2014;15:1129-36. Crossref

11. Centre of Assisted Reproduction and Embryology, The University of Hong Kong–Queen Mary Hospital. How to make an appointment. Available from: https://hkuivf.hku.hk/en/services/fertility-preservation/how-to-make-an-appointment/. Accessed 27 Nov 2024.

12. Ko JK, Lam KK, Cheng HH, et al. Fertility preservation programme in a tertiary-assisted reproduction unit in Hong Kong. Fertil Reprod 2021;3:94-100. Crossref

13. Centre of Assisted Reproduction and Embryology, The University of Hong Kong–Queen Mary Hospital. Fertility preservation. Available from: https://hkuivf.hku.hk/en/services/fertility-preservation/. Accessed 27 Nov 2024.

14. Centre of Assisted Reproduction and Embryology, The University of Hong Kong–Queen Mary Hospital. YouTube Channel. Available from: https://www.youtube.com/@HKUIVF. Accessed 27 Nov 2024.

15. Hong Kong e-Legislation, Hong Kong SAR Government. Cap 561 Human Reproductive Technology Ordinance. Available from: https://www.elegislation.gov.hk/hk/cap561!en-zh-Hant-HK. Accessed 27 Nov 2024.

16. Letourneau JM, Ebbel EE, Katz PP, et al. Pretreatment fertility counseling and fertility preservation improve quality of life in reproductive age women with cancer. Cancer 2012;118:1710-7. Crossref

17. Oktay K, Harvey BE, Partridge AH, et al. Fertility preservation in patients with cancer: ASCO clinical practice guideline update. J Clin Oncol 2018;36:1994-2001. Crossref

18. Blumenfeld Z. Fertility preservation using GnRH agonists: rationale, possible mechanisms, and explanation of controversy. Clin Med Insights Reprod Health 2019;13:1179558119870163. Crossref

19. Zong X, Yu Y, Yang H, et al. Effects of gonadotropin-releasing hormone analogs on ovarian function against chemotherapy-induced gonadotoxic effects in premenopausal women with breast cancer in China: a randomized clinical trial. JAMA Oncol 2022;8:252-8. Crossref

20. Chen H, Xiao L, Li J, Cui L, Huang W. Adjuvant gonadotropin-releasing hormone analogues for the prevention of chemotherapy-induced premature ovarian failure in premenopausal women. Cochrane Database Syst Rev 2019;3:CD008018. Crossref

21. Lambertini M, Ceppi M, Poggio F, et al. Ovarian suppression using luteinizing hormone-releasing hormone agonists during chemotherapy to preserve ovarian function and fertility of breast cancer patients: a meta-analysis of randomized studies. Ann Oncol 2015;26:2408-19. Crossref

22. Demeestere I, Brice P, Peccatori FA, et al. No evidence for the benefit of gonadotropin-releasing hormone agonist in preserving ovarian function and fertility in lymphoma survivors treated with chemotherapy: final long-term report of a prospective randomized trial. J Clin Oncol 2016;34:2568-74. Crossref

23. Moravek MB, Confino R, Lawson AK, et al. Predictors and outcomes in breast cancer patients who did or did not pursue fertility preservation. Breast Cancer Res Treat 2021;186:429-37. Crossref

24. Arecco L, Blondeaux E, Bruzzone M, et al. Safety of fertility preservation techniques before and after anticancer treatments in young women with breast cancer: a systematic review and meta-analysis. Hum Reprod 2022;37:954-68. Crossref

25. Takae S, Lee JR, Mahajan N, et al. Fertility preservation for child and adolescent cancer patients in Asian countries. Front Endocrinol (Lausanne) 2019;10:655. Crossref

26. Chung JP, Chan DY, Song Y, et al. Implementation of ovarian tissue cryopreservation in Hong Kong. Hong Kong Med J 2023;29:121-31. Crossref

Due to the increasing incidence of osteoarthritis associated with the ageing global population, total knee arthroplasty (TKA) has become one of the most commonly performed orthopaedic procedures worldwide. The most common approach to determine causes and risk factors involves analysing readmission episodes among TKA patients within 30 days post-surgery.1 2 The 30-day readmission rate provides insight into the prevalence of postoperative complications, whereas the length of stay after readmission reflects the severity of those complications. A review of readmission causes is needed to assess the quality of hospital care and determine the adequacy of patient education (eg, wound management).3 An understanding of the 30-day readmission rate, causes, and risk factors can help hospitals improve clinical guidelines, reduce medical and surgical complications,4 and reduce the financial burden of treatment for these complications.5

Although multiple studies worldwide have adequately explored the 30-day readmission causes, rate, and length of stay among TKA patients,6 7 revealing important clinical insights, no such studies have been conducted in Hong Kong. The current study aimed to investigate the 30-day readmission rate, causes, and risk factors among TKA patients in the city.

This retrospective study included all patients who underwent TKA at our local university-affiliated hospital and were readmitted through an Accident and Emergency Department (AED) between 2001 and 2020. It evaluated the epidemiological characteristics, readmission causes, and preoperative co-morbidities of TKA patients.

We utilised data from the Clinical Data Analysis and Reporting System (CDARS), a well-established platform developed by the Hospital Authority (HA). The CDARS contains patient data, such as laboratory reports and radiological images; it covers all outpatients and inpatients at 43 public hospitals and institutions across seven service clusters in Hong Kong. Records in the CDARS include the details of patients with unplanned 30-day readmission to the AED of an HA hospital from either their homes or rehabilitation facilities, along with their discharge information. This platform is extensively used by research teams across Hong Kong.8 We obtained a list of TKA patients who underwent surgery at the study hospital and were readmitted to an HA hospital within 30 days. We matched these patient names with their corresponding electronic patient records to determine the reasons for readmission.

For patients who experienced 30-day readmission, both the records in the CDARS and electronic patient records were reviewed. For patients who did not require 30-day readmission, only CDARS records were reviewed. Medication records (ie, dispensing dates, dosages, and durations) were extracted from CDARS records to identify co-morbidities (online supplementary Appendix). All patient data were de-identified.

Based on factors described by Roger et al,6 we classified reasons for readmission into the following categories: orthopaedics-related, surgery-related, gastrointestinal-related, urological-related, neurological-related, cardiac-related, respiratory-related, renal-related, medication-related, and others. Orthopaedic specialists performed the classification to determine the cause of readmission.

The inclusion criteria were a recent history of TKA at our institution, readmission through the AED of an HA hospital, and inpatient admission. The exclusion criteria were a history of knee surgery, incomplete clinical assessment data, and/or orthopaedic tumours in the knee (for paediatric patients only).

Analyses of readmission cause, number, and rate, as well as organ dysfunction episodes, were episode-based. The analysis of risk factors for readmission was patient-based. Risk factors/co-morbidities were identified based on medications prescribed to the patients. If a patient received antihypertensive medication, that patient was assumed to have hypertension.

Data analysis was performed using R (R Foundation for Statistical Computing, Vienna, Austria) and R Studio software. All statistical tests were two-sided, and a 5% significance threshold was applied. The investigators and their research assistants were responsible for data collection and had access to the source data and study records. To evaluate categorical variables, Chi squared tests and/or Fisher’s exact tests were conducted, depending on the observed frequencies. To evaluate continuous variables, the Kruskal–Wallis test was used.

In total, 3878 records were initially reviewed; of these, 43 were excluded due to the presence of tumours (ie, osteosarcoma in the distal tibia), three were excluded due to incorrect data entry for revision surgery, and five were excluded because they constituted duplicate entries for the same readmission episode (online supplementary Fig 1).

Of the 3827 valid patient records, 2855 were included in the initial analysis after removal of duplicate records for 972 patients who underwent two unilateral TKAs during different admission episodes. Of the 3827 patients, 2815 (73.6%) were women and 1012 (26.4%) were men. The mean ages were 71.11 years for TKA patients who did not experience readmission and 73.10 years for TKA patients who experienced readmission (Table 16 9 10 11 12). The mean postoperative length of stay (±standard deviation) was 6.85±6.19 days (Table 2). Thus, the readmission rate at our institution was 3.4%, similar to rates reported worldwide (Table 1).6 7 13 There was an increase in the 30-day readmission rate between 2001 and 2020 (Table 2). The number of TKAs performed in our institution increased from 2001 to 2014 and remained consistently high (>200 TKAs annually except in 2020) [online supplementary Fig 2], in line with published literature.14

In total, 130 patients with valid readmission records were analysed to identify causes and risk factors (online supplementary Fig 1). Of these patients, 90 (69.2%) were women and 40 (30.8%) were men. The median length of stay after readmission was 2 days (interquartile range=1.25-5) and the mean time between surgery and readmission was 22 days (Table 3).

Unilateral knee pain (33.1%), unilateral knee swelling (26.2%), and gastrointestinal-related conditions (8.5%) were the most common causes of readmission (Table 4). Hypertension (67.7%) and diabetes mellitus (22.7% before March 2017 from which our institution modified the preoperative management pathway) were the most common co-morbidities among readmitted patients. Additionally, hypertension (odds ratio [OR]=2.08; P<0.001) and age ≥80 years (OR=1.63; P=0.01) were identified as significant risk factors for readmission (Table 5 and online supplementary Table 1). Patients who underwent bilateral TKA had a 58% lower risk of readmission (Table 6), possibly because they had better health condition before surgery and received more rigorous preoperative screening for high-risk co-morbidities.

Our results suggest that there is a substantial rate of readmission due to pain and swelling among TKA patients in Hong Kong, which is higher than the rates in previous studies (Table 1).6 9 12 Hospital resources should be reviewed to determine whether these patients require admission because most readmitted patients have non-severe conditions. To reduce the unnecessary allocation of clinical resources to non-severe cases, alternatives such as designated nurse clinics15 and patient consultation hotlines can provide medical advice for managing minor conditions (eg, pain and swelling) at home. These measures can reduce the workload of orthopaedic surgeons and improve postoperative follow-up care.

Hypertension and age ≥80 years were significant risk factors for readmission. The mean age of TKA patients in Hong Kong was higher than the mean ages of TKA patients in similar studies worldwide (Table 1)6 9 10 11 12; this difference aligns with the fact that Hong Kong has the longest life expectancy globally (mean age of 85.16 years in 2022).16

The increased risk of readmission with old age, consistent with findings in a previous study,17 may be related to the greater likelihood for older individuals to visit the AED for non-orthopaedic issues. In contrast, patients who underwent simultaneous bilateral TKA had a lower risk of readmission (OR=0.42; P=0.005) [Table 6].

At our institution, patients who underwent simultaneous bilateral TKA were aged <75 years and had no clinically significant cardiovascular co-morbidities (eg, stroke). Furthermore, in March 2017, our institution introduced routine glycated haemoglobin screening to identify diabetic and prediabetic patients, with the goal of minimising postoperative complications. Diabetes mellitus is known to increase the risk of periprosthetic joint infection after surgery.18 Patients with elevated glycated haemoglobin levels were referred to endocrinologists for better management of diabetes mellitus prior to TKA, thereby decreasing the OR for readmission from 1.24 to 0.74 (Table 5 and online supplementary Table 1). Considering that a substantial number of patients with the aforementioned co-morbidities exhibit a higher risk of readmission, the perioperative protocol could be improved. Suggested changes could include better coordination with each patient’s family medicine specialists or general practitioners, who usually have a better understanding of the patient’s underlying medical conditions, to develop effective preoperative and postoperative care plans.

Overall, 4.6% (n=6) of the patients were readmitted primarily due to falls (Table 4). This finding highlights the need for enhanced nursing support and education to prevent post-surgical falls among TKA patients. Furthermore, stronger occupational therapy and household aid programmes can help prevent falls at home and improve patient rehabilitation. Another 4.6% of patients were readmitted due to the adverse drug effects (Table 4 and online supplementary Table 2), particularly from tramadol/codeine/morphine and related medications; symptoms included vomiting and constipation. These results indicate a need for cautious painkiller prescribing to prevent future medication-related readmissions.

To the best of our knowledge, this is the first study on the readmission rate, causes, and risk factors among TKA patients within 30 days post-surgery. It also compared data collected in Hong Kong with results from other studies, yielding insights for local orthopaedic surgeons who seek to improve suboptimal surgical outcomes.

Notably, there were some limitations. Patient data in the CDARS may be incomplete because some doctors might have omitted the International Classification of Diseases, Ninth Revision codes for certain co-morbidities. To mitigate this issue, prescribed drugs were used to identify patients’ co-morbidities. However, this approach may have missed some patients with co-morbidities and no associated medication records.

This study showed that older TKA patients with hypertension were more likely to be readmitted through the AED within 30 days post-surgery. The most common reasons for readmission were pain, swelling, and gastrointestinal-related symptoms. To reduce readmissions, hospitals should place greater emphasis on pain and wound management for TKA patients. Furthermore, patient education efforts should be strengthened to increase awareness of pain and wound management.

Concept or design: OWK Tsui, PK Chan.
Acquisition of data: OWK Tsui, PK Chan, JHY Leung.
Analysis or interpretation of data: OWK Tsui, PK Chan, JHY Leung.
Drafting of the manuscript: OWK Tsui.
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.

All authors have disclosed no conflicts of interest.

The research was presented at 42nd Annual Congress of the Hong Kong Orthopaedic Association, 5 November 2022, 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 Institutional Review Board of The University of Hong Kong/Hospital Authority Hong Kong West Cluster, Hong Kong (Ref No.: UW-22-313). The requirement for patient consent was waived by the Board 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. Howie CM, Mears SC, Barnes CL, Stambough JB. Readmission, complication, and disposition calculators in total joint arthroplasty: a systemic review. J Arthroplasty 2021;36:1823-31. Crossref

2. D’Apuzzo M, Westrich G, Hidaka C, Jung Pan T, Lyman S. All-cause versus complication-specific readmission following total knee arthroplasty. J Bone Joint Surg Am 2017;99:1093-103. Crossref

3. Chambers MC, El-Othmani MM, Anoushiravani AA, Sayeed Z, Saleh KJ. Reducing 30-day readmission after joint replacement. Orthop Clin North Am 2016;47:673-80. Crossref

4. Bosco JA 3rd, Karkenny AJ, Hutzler LH, Slover JD, Iorio R. Cost burden of 30-day readmissions following Medicare total hip and knee arthroplasty. J Arthroplasty 2014;29:903-5. Crossref

5. Gould D, Dowsey MM, Spelman T, et al. Patient-related risk factors for unplanned 30-day hospital readmission following primary and revision total knee arthroplasty: a systematic review and meta-analysis. J Clin Med 2021;10:134. Crossref

6. Roger C, Debuyzer E, Dehl M, et al. Factors associated with hospital stay length, discharge destination, and 30-day readmission rate after primary hip or knee arthroplasty: retrospective cohort study. Orthop Traumatol Surg Res 2019;105:949-55. Crossref

7. Bovonratwet P, Shen TS, Ast MP, Mayman DJ, Haas SB, Su EP. Reasons and risk factors for 30-day readmission after outpatient total knee arthroplasty: a review of 3015 cases. J Arthroplasty 2020;35:2451-7. Crossref

8. Sing CW, Woo YC, Lee AC, et al. Validity of major osteoporotic fracture diagnosis codes in the Clinical Data Analysis and Reporting System in Hong Kong. Pharmacoepidemiol Drug Saf 2017;26:973-6. Crossref

9. Phruetthiphat OA, Otero JE, Zampogna B, Vasta S, Gao Y, Callaghan JJ. Predictors for readmission following primary total hip and total knee arthroplasty. J Orthop Surg (Hong Kong) 2020;28:2309499020959160. Crossref

10. Ross TD, Dvorani E, Saskin R, Khoshbin A, Atrey A, Ward SE. Temporal trends and predictors of thirty-day readmissions and emergency department visits following total knee arthroplasty in Ontario between 2003 and 2016. J Arthroplasty 2020;35:364-70. Crossref

11. Kurtz SM, Lau EC, Ong KL, Adler EM, Kolisek FR, Manley MT. Which hospital and clinical factors drive 30- and 90-day readmission after TKA? J Arthroplasty 2016;31:2099-107. Crossref

12. Ali AM, Loeffler MD, Aylin P, Bottle A. Predictors of 30-day readmission after total knee arthroplasty: analysis of 566,323 procedures in the United Kingdom. J Arthroplasty 2019;34:242-8.e1. Crossref

13. Urish KL, Qin Y, Li BY, et al. Predictors and cost of readmission in total knee arthroplasty. J Arthroplasty 2018;33:2759-63. Crossref

14. Yan CH, Chiu KY, Ng FY. Total knee arthroplasty for primary knee osteoarthritis: changing pattern over the past 10 years. Hong Kong Med J 2011;17:20-5.

15. Fan JC, Lo CK, Kwok CK, Fung KY. Nurse-led orthopaedic clinic in total joint replacement. Hong Kong Med J 2014;20:511-8. Crossref

16. MacroTrends. Hong Kong life expectancy 1950-2024. Available from: https://www.macrotrends.net/countries/HKG/hong-kong/life-expectancy. Accessed 28 Nov 2024.

17. Cheung A, Fu H, Cheung MH, et al. How well do elderly patients do after total knee arthroplasty in the era of fasttrack surgery? Arthroplasty 2020;2:16. Crossref

18. Chan VW, Chan PK, Woo YC, et al. Universal haemoglobin A1c screening reveals high prevalence of dysglycaemia in patients undergoing total knee arthroplasty. Hong Kong Med J 2020;26:304-10. Crossref

The International Study of Asthma and Allergies in Childhood (ISAAC) is considered a landmark international investigation of the global burden of three major atopic diseases—asthma, allergic rhinitis, and eczema1 2 3 4—utilising standardised methodology to enable accurate estimation of prevalence trends in allergic diseases. Phase One examined the prevalence and severity of the three diseases in 1994-1995, Phase Two investigated the corresponding risk factors, and Phase Three constituted a follow-up study examining global prevalence and severity trends in 2001-2003.1 2 4 The Global Asthma Network updated global burden information using data from 27 Global Asthma Network Centres in 14 countries in 2017-2020,3 but Hong Kong data were excluded from that study.

Asthma remains a major medical burden responsible for nearly 500 000 deaths in 20175 and substantial direct, indirect, and intangible costs,6 ranging from mortality and hospitalisations to school- or work-related loss (in the form of absenteeism from school or loss of working days), and quality of life impairments. Thus, accurate and current asthma prevalence data are essential for guiding public health initiatives and formulating healthcare policies. Recent estimates suggest global asthma prevalences of 9.1% in children, 11.0% in adolescents, and 6.6% in adults7; asthma is often cited as the most common chronic disease in children.6

Trends regarding the three major atopic diseases have considerably varied among countries and regions since 1994-1995. Asthma has shown the greatest variation, peaking in some countries but continuing to increase in others, especially developing countries.2 3 4 5 6 8 9 10 11 12 13 14 15 16 Most studies of asthma prevalence were conducted before the coronavirus disease 2019 (COVID-19) pandemic, and statistics considerably differ even within the same region. In 2019, a large-scale study examining the global adult and paediatric prevalences of asthma in >200 territories showed a 24% decrease in age-standardised point prevalence in most countries5; however, increases were observed in Oman, Saudi Arabia, and Vietnam.5 This discrepancy highlights the need for further investigation of the risk and protective factors for asthma in modern societies.

Hong Kong is an urban-centric port city in southeast China with a unique blend of Chinese and international influences, which have led to trends that considerably differ from those of neighbouring regions.17 Cultural practices17 18 19 and urbanisation (eg, cooking methods, joss stick burning, and high population density in urban areas)9 17 20 21 increase the complexity involved in estimating asthma prevalence within Hong Kong based on data from neighbouring regions. This study presents contemporary epidemiological data regarding the prevalence and severity of asthma and wheezing among children in Hong Kong, which provides an update to the results of the ISAAC conducted in 1994-1995 and 2001-2003.

This cross-sectional study, based on the ISAAC study protocol,2 was performed using traditional Chinese versions of the validated ISSAC measurements.22

In total, 2148 children aged 6 to 7 years or 13 to 14 years residing in Hong Kong were recruited from primary and secondary schools between September 2020 and August 2021. Inclusion criteria for children were enrolment in Primary 1-2 (Grades 1-2) or Secondary 2-3 (Grades 8-9) in Hong Kong and parental provision of informed consent to participate. The primary caregiver of the student (for the primary school group) or the student himself/herself (for the secondary school group) was asked to complete questionnaires in accordance with the ISAAC protocol. Only written questionnaires were used, considering reports of high agreement with their video counterparts.23 24 Inclusion criteria for parents were: (1) being either the father or mother of the child, (2) bearing primary responsibility for the child’s care for ≥6 months, and (3) provision of informed consent to participate. Exclusion criteria for parents and children were the presence of a learning disability or organic disorder that would impair the ability to understand and respond to the questionnaires, and inability to understand Chinese. Additionally, children attending Special Educational Needs schools were excluded.

All schools in Hong Kong (n=900) were invited by phone and mail to participate in the survey. Contact persons were identified in each school. The investigators made at least three attempts per school (by phone, mail and/or email) to obtain responses from participating students. A target sample size of 1000 to 3000 participants was established, in accordance with the ISAAC study protocol.2

Questions were adopted from the ISAAC study protocol, and disease definitions were based on the original ISAAC Phase Three questionnaire and handbook.2 ‘Current severe asthma’ was defined as affirmative responses to one or more of the following items in the past 12 months: (1) ≥4 wheezing episodes, (2) woken from sleep by wheezing ≥1 night per week, or (3) limitation of speech during wheezing. Other questions and their definitions adhered to the standardised ISAAC Coding and Data Transfer Manual2 22 to ensure comparability with previous studies utilising the ISAAC protocol.

Descriptive statistics, including means and standard deviations, were used for continuous variables; frequency distributions were used for categorical variables. Sex differences within each age-group were assessed using Chi squared and independent t tests. Demographic information was entered into hierarchical logistic regression models to identify potential predictive factors associated with asthma severity. Annual change in prevalence was calculated via division of the prevalence by the number of years between surveys. Logistic regression modelling was conducted to identify risk factors associated with asthma severity. Potential risk factors included parent gender, parent education, and household monthly income. All analyses were performed using SPSS software (Windows version 26.0; IBM Corp, Armonk [NY], United States). P values <0.05 were considered statistically significant.

Nineteen primary schools and 25 secondary schools accepted the invitation to participate, representing 3748 and 5771 eligible students in the respective groups. Overall, 1970 (52.6%) questionnaires were returned for the primary school group, whereas 1968 (34.1%) questionnaires were returned for the secondary school group; 1100 and 1048 questionnaires were analysed in the respective groups after exclusion due to invalid consent, incomplete data, or ineligibility (Fig). Boys comprised 62.51% and 42.64% of the primary and secondary school groups, respectively (Tables 1 and 2). The mean ages were 7.02 years (standard deviation [SD]=0.76) in the primary school group and 14.09 years (SD=0.89) in the secondary school group.

The estimated prevalences of current wheezing were 6.19% and 4.97% in the primary school and secondary school groups, respectively. The respective prevalences of exercise-induced wheezing and night cough were 7.46% and 16.74% in the primary school group and 17.88% and 10.61% in the secondary school group (Tables 1 and 2).

The prevalences of asthma ever were 5.55% and 6.12% in the primary school and secondary school groups, respectively; significantly more boys reported asthma ever in the secondary school group. The prevalences of bronchial hypersensitivity ever were 33.03% and 11.09% in the primary and secondary school groups, respectively (Tables 1, 2, 3). In total, 85.25% and 78.13% of primary and secondary school participants with asthma exhibited comorbid eczema and/or allergic rhinitis, respectively (Table 4)

Indicators of severe asthma were also examined. The incidences of ≥4 wheezing episodes in the past year were 2.18% (boys vs girls: 3.20% vs 0.49%) in the primary school group and 1.91% (boys vs girls: 2.24% vs 1.67%) in the secondary school group. Severe wheezing that limited speech in the past year was reported by 0.64% and 1.43% of participants in the primary and secondary school groups, respectively. Waking from sleep due to wheezing in the past year occurred in 2.82% and 1.24% of participants in the primary and secondary school groups, respectively. Hospitalisation due to shortness of breath was reported by 7.37% (boys vs girls: 8.30% vs 5.83%) and 4.30% (boys vs girls: 4.93% vs 3.83%) of participants in the primary and secondary school groups, respectively (Tables 1 and 2). The mean ages of wheezing onset in the primary and secondary school groups were 2.37 years (SD=1.52) and 7.16 years (SD=4.38), respectively.

Compared with the previous Hong Kong ISAAC data,1 25 26 27 the prevalence of wheezing ever increased from 16.80%26 in 1994-1995 to 20.38% in the primary school group but decreased from 20%25 to 12.05% in the secondary school group. Conversely, the prevalence of asthma ever decreased from 7.80%26 to 5.55% in the primary school group and from 11%26 to 6.12% in the secondary school group. The prevalence of wheezing in the past year decreased from 9.20% in 1994-199526 and 9.40% in 2000-200127 to 6.19% in our study (2020-2021) in the primary school group; it decreased from 12% in 1994-1995 to 4.97% in our study in the secondary school group. The annual changes in wheezing prevalence were -0.16% and -0.27% in the primary and secondary school groups, respectively. The aforementioned indicators of severe asthma, including ≥4 wheezing episodes in the past year and severe wheezing that limited speech in the past year, decreased in prevalence compared with the 1994-1995 figures. However, the prevalence of waking from sleep due to wheezing in the past year decreased in the primary group and increased in the secondary school group (Table 5).

Overall, both the prevalence and severity of asthma declined compared with the 1994-1995 data. We observed a statistically significant male predominance for wheezing ever (P=0.003), current wheezing (P=0.012), and bronchial hypersensitivity diagnosis (P=0.002) in the primary school group (Table 1), as well as for asthma ever in the secondary school group (P=0.005) [Table 2].

In the logistic regression model, demographic characteristics, including parent gender, household income, and parent education, were not significantly predictive of wheezing episodes in the past year. Hierarchical logistic regression using demographic characteristics along with concomitant current eczema and rhinitis also did not show significantly predictive effects for wheezing episodes in the past year.

This is the first study since 2001 to investigate trends in asthma prevalence and severity among school children in Hong Kong. The global variability in these trends is complicated by the lack of consensus regarding exact definitions of asthma entities, the heterogeneity of the disease itself, changes in community awareness, the absence of a ‘gold standard’ diagnostic test, and the non-specific nature of symptoms shared with other diagnoses.13 This study showed a decrease in asthma prevalence, consistent with findings from the neighbouring region of Taiwan.11

The 2020 study by Asher et al4 investigating international symptom trends showed annual current asthma prevalence increases of 0.06% in the secondary school group (from 13.2% to 13.7%) and 0.13% in the primary school group (from 11.1% to 11.6%), and increases in the prevalence of asthma ever by 0.18% and 0.28% in the primary and secondary school groups, respectively. However, trends varied across regions; in general, asthma prevalence decreased in higher-income regions but increased in low- and middle-income regions. Considering that Hong Kong exhibits higher levels of income and gross domestic product per capita, it is unsurprising that the local asthma prevalences showed a decreasing trend.

A decrease in asthma severity was also observed in most regions, reflected by the three indicators of severe asthma in our study and parameters such as asthma-related hospital admissions and mortality. These findings suggest that, regardless of overall asthma trends, milder forms of asthma have become more prevalent in recent years.7 28 Our study revealed similar trends in terms of fewer severe asthmatic exacerbations and flares. However, according to the ISAAC protocol, centres with 1000 to 2999 participants are considered appropriate for comparisons of prevalence, but not severity, with centres in other regions.2

Safiri et al5 identified the highest asthma prevalence among 5- to 9-year-olds, which then decreased and remained stable until adulthood. A proportion of young children with wheezing may have ‘transient wheezing’ that does not progress to asthma.4 This notion is consistent with our finding of less frequent symptoms in the past year among older children.5 The prevailing view is that approximately 50% of preschool children with wheezing will progress to asthma by the time of primary school entry.4 The prevalences of 12.05% and 6.12% for wheezing ever and asthma ever observed in our secondary school group are consistent with this view.

Our results showed a male predominance for asthma and wheezing in the primary school group, consistent with existing literature.29 Various mechanisms have been proposed to explain the post-pubertal shift towards female predominance due to sex hormone changes, genetic and epigenetic differences, co-morbidities, and socio-economic factors.14 The significantly older age of wheezing onset reported in the secondary school group, along with the lower prevalence of wheezing ever in the secondary school group, may be attributable to recall bias due to the use of questionnaires, a limitation noted in other studies.30

Co-morbidities with other atopic diseases were common among our asthma participants; most participants in both groups had concomitant atopic dermatitis, allergic rhinitis, and asthma, consistent with previous reports.8 16 Otherwise, our study did not identify relevant demographic characteristics or risk factors through statistical analysis that could predict wheezing episodes. Many studies have investigated various factors potentially associated with asthma diagnosis, wheezing, exercise-induced symptoms, and nocturnal cough. These factors include, but are not limited to, family history, recurrent respiratory infections, early-life severe respiratory syncytial virus infection, exposure to cigarette smoke, exposure to pets, incense burning, maternal education level, sex, race, vaccination rates, humidity, air pollution index (particulate matter) and exposure (particularly nitrogen dioxide and sulphur dioxide), exposure to indoor mould, farm residence, exposure to indoor endotoxins, socio-economic status, obesity, open fire cooking, nutritional levels, neonatal antibiotic use, delivery mode, urban living environment, psychosocial environment (including maternal stress), current paracetamol use, maternal antibiotic use during pregnancy, maternal vitamin D consumption during pregnancy, maternal weight gain during pregnancy, maternal paracetamol use during pregnancy, proton-pump inhibitor and H2-receptor antagonist use, and new immigrant status.2 3 4 5 6 8 9 10 11 12 13 14 15 29 The next phase of our study will examine these predictive or protective factors; it will also explore risk factors unique to our population, including cultural perceptions,18 feeding and weaning practices,19 31 joss stick burning,15 and exposure to traditional Chinese herbal medicine.32 33

Due to its survey approach, this study has limitations of recall bias and cross-sectional design. The estimations of symptom prevalence also lack objective confirmation through medical assessments and objective tests, which could lead to over- or underestimation of the asthmatic population; this aspect is further complicated by diagnostic discrepancies among medical professionals across regions and time periods.4 34 Although our study population exceeded 1000 students per group, it was smaller than the original Hong Kong ISAAC studies, which included >300026 and 400025 participants, respectively. The paediatric population size has also significantly changed, according to the Hong Kong Population Census data: 86 000 13- to 14-year-olds in 1994 compared with 61 800 in 2021.35 The male-to-female ratio of 1.67 and 0.74 in the primary and secondary school groups, respectively, differed from the 2021 census figures (1.063 and 1.067, respectively)35; therefore, our findings may not reflect the true prevalence and severity of asthma in the Hong Kong paediatric population, which represents a key limitation of the study. Although the ISAAC protocol acknowledges the impracticality of video questionnaires due to logistic or technical factors,2 exclusive use of the written questionnaire may have impacted our results; notably, a study has shown that video and written questionnaires are comparable.23 The relatively large population size and the use of validated and standardised questionnaires may partially mitigate these limitations.

In this study, response rates were lower than in the previous ISAAC studies (where rates >80% for most centres),36 likely due to ‘survey fatigue’ and challenges in motivating participants during the COVID-19 pandemic. Similarly low response rates were observed in studies requesting parental completion of home questionnaires.37 Although no significant differences in airway disease or symptom prevalences have been identified between non-responders and responders,38 the lower response rates may have introduced selection bias into our study.

The increased prevalence of wheezing ever in the primary school group, decreased prevalence of wheezing ever in the secondary school group, decreased prevalences of asthma ever in both groups, and decreased symptom severity all reflect enhanced awareness and modified diagnostic practices among the general population and medical professionals.4 11 13 Doctors are less likely to classify patients as ‘asthmatic’ without collecting a thorough clinical history and performing diagnostic testing, leading to a larger difference in the prevalences of wheezing and asthma.11 13

The local COVID-19 policy, commonly known as the mask mandate, along with social isolation, increased awareness of infection control, changes in drug compliance, and enhanced hygiene practices, may have significantly reduced triggers for infectious and allergic airway diseases.39 40 However, mechanisms linking COVID-19 to severe asthma risk have also been proposed.41 These policies and social practices may explain the overall decreases in asthma prevalence and severity observed in our study. The timing of the study coincided with the COVID-19 pandemic, where the modified local health practices may have influenced trends concerning current asthma and airway symptoms.

This study provides an essential update regarding the prevalences of asthma and other respiratory symptoms among school children in Hong Kong. Our findings indicate overall decreasing trends in asthma severity and prevalence. A follow-up study will explore the protective and risk factors contributing to these trends.

Concept or design: JWCH Cheng, YP Tsang, YY Lam, AKY Chu, CHY Chan, YL Fung, PSY Chau, DCK Luk.
Acquisition of data: All authors.
Analysis or interpretation of data: All authors.
Drafting of the manuscript: JWCH Cheng, CSY Ng.
Critical revision of the manuscript for important intellectual content: JWCH Cheng, CSY Ng, CHY Chan, YL Fung.

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 all the children and their families for their participation in this research.

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 Human Research Ethics Committee of The University of Hong Kong, Hong Kong (Ref No.: EA2002007). All study participants provided written consent for publication of their data which were de-identified in this article.

1. Asher MI, Montefort S, Björkstén B, et al. Worldwide time trends in the prevalence of symptoms of asthma, allergic rhinoconjunctivitis, and eczema in childhood: ISAAC Phases One and Three repeat multicountry cross-sectional surveys. Lancet 2006;368:733-43. Crossref

2. Asher MI, Keil U, Anderson HR, et al. International Study of Asthma and Allergies in Childhood (ISAAC): rationale and methods. Eur Respir J 1995;8:483-91. Crossref

3. Asher MI, Rutter CE, Bissell K, et al. Worldwide trends in the burden of asthma symptoms in school-aged children: Global Asthma Network Phase I cross-sectional study. Lancet 2021;398:1569-80. Crossref

4. Asher MI, García-Marcos L, Pearce NE, Strachan DP. Trends in worldwide asthma prevalence. Eur Respir J 2020;56:2002094. Crossref

5. Safiri S, Carson-Chahhoud K, Karamzad N, et al. Prevalence, deaths, and disability-adjusted life-years due to asthma and its attributable risk factors in 204 countries and territories, 1990-2019. Chest 2022;161:318-29. Crossref

6. Ferrante G, La Grutta S. The burden of pediatric asthma. Front Pediatr 2018;6:186. Crossref

7. The Global Asthma Report 2022 [editorial]. Int J Tuberc Lung Dis 2022;26(Suppl 1):1-104 Crossref

8. Kansen HM, Le TM, Uiterwaal C, et al. Prevalence and predictors of uncontrolled asthma in children referred for asthma and other atopic diseases. J Asthma Allergy 2020;13:67-75. Crossref

9. Rodriguez A, Brickley E, Rodrigues L, Normansell RA, Barreto M, Cooper PJ. Urbanisation and asthma in low-income and middle-income countries: a systematic review of the urban–rural differences in asthma prevalence. Thorax 2019;74:1020-30.Crossref

10. Castro-Rodriguez JA, Forno E, Rodriguez-Martinez CE, Celedón JC. Risk and protective factors for childhood asthma: what is the evidence? J Allergy Clin Immunol Pract 2016;4:1111-22. Crossref

11. Chen WY, Lin CW, Lee J, Chen PS, Tsai HJ, Wang JY. Decreasing ten-year (2008-2018) trends of the prevalence of childhood asthma and air pollution in Southern Taiwan. World Allergy Organ J 2021;14:100538. Crossref

12. Akinbami LJ, Simon AE, Rossen LM. Changing trends in asthma prevalence among children. Pediatrics 2016;137:1-7. Crossref

13. Sears MR. Trends in the prevalence of asthma. Chest 2014;145:219-25. Crossref

14. Chowdhury NU, Guntur VP, Newcomb DC, Wechsler ME. Sex and gender in asthma. Eur Respir Rev 2021;30:210067. Crossref

15. Lin TC, Krishnaswamy G, Chi DS. Incense smoke: clinical, structural and molecular effects on airway disease. Clin Mol Allergy 2008;6:3. Crossref

16. Arasi S, Porcaro F, Cutrera R, Fiocchi AG. Severe asthma and allergy: a pediatric perspective. Front Pediatr 2019;7:28. Crossref

17. Wong GW, Ko FW, Hui DS, et al. Factors associated with difference in prevalence of asthma in children from three cities in China: multicentre epidemiological survey. BMJ 2004;329:486. Crossref

18. Leo HL, Wang A, Gong M, Clark N. Cultural perceptions on identification and treatment of childhood asthma in the US and China. J Allergy Clin Immunol 2006;117 Suppl:S53. Crossref

19. Mihrshahi S, Ampon R, Webb K, et al. The association between infant feeding practices and subsequent atopy among children with a family history of asthma. Clin Exp Allergy 2007;37:671-9. Crossref

20. Wong GW, Brunekreef B, Ellwood P, et al. Cooking fuels and prevalence of asthma: a global analysis of phase three of the International Study of Asthma and Allergies in Childhood (ISAAC). Lancet Respir Med 2013;1:386-94. Crossref

21. Asher MI. Urbanisation, asthma and allergies. Thorax 2011;66:1025-6. Crossref

22. Chan HH, Pei A, Van Krevel C, Wong GW, Lai CK. Validation of the Chinese translated version of ISAAC core questions for atopic eczema. Clin Exp Allergy 2001;31:903-7. Crossref

23. Lai CK, Chan JK, Chan A, et al. Comparison of the ISAAC video questionnaire (AVQ3.0) with the ISAAC written questionnaire for estimating asthma associated with bronchial hyperreactivity. Clin Exp Allergy 1997;27:540-5. Crossref

24. Becerir T, Akcay A, Duksal F, Ergin A, Becerir C, Guler N. Prevalence of asthma, local risk factors and agreement between written and video questionnaires among Turkish adolescents. Allergol Immunopathol (Madr) 2014;42:594-602. Crossref

25. Leung R, Wong G, Lau J, et al. Prevalence of asthma and allergy in Hong Kong schoolchildren: an ISAAC study. Eur Respir J 1997;10:354-60. Crossref

26. Lau YL, Karlberg J. Prevalence and risk factors of childhood asthma, rhinitis and eczema in Hong Kong. J Paediatr Child Health 1998;34:47-52. Crossref

27. Lee SL, Wong W, Lau YL. Increasing prevalence of allergic rhinitis but not asthma among children in Hong Kong from 1995 to 2001 (Phase 3 International Study of Asthma and Allergies in Childhood). Pediatr Allergy Immunol 2004;15:72-8. Crossref

28. Ebmeier S, Thayabaran D, Braithwaite I, Bénamara C, Weatherall M, Beasley R. Trends in international asthma mortality: analysis of data from the WHO Mortality Database from 46 countries (1993-2012). Lancet 2017;390:935-45. Crossref

29. Estela DB, Arturo B, Nayely RN, et al. Have asthma symptoms in Mexico changed in the past 15 years? Time trends from the International Study of Asthma and Allergies in Childhood to the Global Asthma Network. Allergol Immunopathol (Madr) 2021;49:1-10. Crossref

30. Strachan DP. The prevalence and natural history of wheezing in early childhood. J R Coll Gen Pract 1985;35:182-4.

31. Yung J, Yuen JW, Ou Y, Loke AY. Factors associated with atopy in toddlers: a case-control study. Int J Environ Res Public Health 2015;12:2501-20. Crossref

32. Li XM. Traditional Chinese herbal remedies for asthma and food allergy. J Allergy Clin Immunol 2007;120:25-31. Crossref

33. Li XM, Brown L. Efficacy and mechanisms of action of traditional Chinese medicines for treating asthma and allergy. J Allergy Clin Immunol 2009;123:297-306; quiz 307-8. Crossref

34. Hederos CA, Hasselgren M, Hedlin G, Bornehag CG. Comparison of clinically diagnosed asthma with parental assessment of children’s asthma in a questionnaire. Pediatr Allergy Immunol 2007;18:135-41. Crossref

35. Census and Statistics Department, Hong Kong SAR Government. Population estimates. Table 110-01002: Population by sex and age. Available from: https://www.censtatd.gov.hk/en/web_table.html?id=110-01002#. Accessed 14 Oct 2024.

36. Worldwide variations in the prevalence of asthma symptoms: the International Study of Asthma and Allergies in Childhood (ISAAC). Eur Respir J 1998;12:315-35. Crossref

37. Ellwood P, Ellwood E, Rutter C, et al. Global Asthma Network Phase I surveillance: geographical coverage and response rates. J Clin Med 2020;9:3688. Crossref

38. Rönmark EP, Ekerljung L, Lötvall J, Torén K, Rönmark E, Lundbäck B. Large scale questionnaire survey on respiratory health in Sweden: effects of late- and nonresponse. Respir Med 2009;103:1807-15. Crossref

39. Izquierdo-Domínguez A, Rojas-Lechuga MJ, Alobid I. Management of allergic diseases during COVID-19 outbreak. Curr Allergy Asthma Rep 2021;21:8. Crossref

40. Carr TF, Kraft M. Asthma and atopy in COVID-19: 2021 updates. J Allergy Clin Immunol 2022;149:562-4. Crossref

41. Hosoki K, Chakraborty A, Sur S. Molecular mechanisms and epidemiology of COVID-19 from an allergist’s perspective. J Allergy Clin Immunol 2020;146:285-99. Crossref

Atopic eczema (AE) is a common childhood skin disease associated with pruritus and sleep disturbance.1 2 3 4 5 Childhood AE can substantially influence quality of life (QOL) among affected children and their parents. The extent of QOL impairment is often correlated with eczema severity, skin dehydration, and staphylococcal infection.6 Additionally, many affected children and their parents develop depression, anxiety, and stress symptoms.1 These mental health issues are correlated with disease severity, impaired QOL, and therapeutic non-compliance.1 7 8 A study using the 42-item Depression, Anxiety, and Stress Scales (DASS-42) found that depression, anxiety, and stress symptoms were present in 21%, 33%, and 23% of Hong Kong adolescent patients with AE, respectively.1 These psychological symptoms were significantly correlated with poor QOL.1 A study of Japanese children showed eczema severity was associated with mental health.9 Furthermore, a retrospective, cross-sectional population-based survey of childhood eczema in the United States revealed that increased eczema severity was associated with a higher risk of mental disorders.10 School-aged children with moderate and severe AE have a higher risk of psychosocial problems that can influence their quality of sleep and cognitive development.11 Emotion, attention, interpersonal relationships, and conduct can also be affected by AE.10 12 Moreover, parents are often unaware of potential psychosocial health issues in their children with eczema.13 Most affected children and their parents do not receive appropriate psychological help and support; they also exhibit low symptom recognition. The impacts of childhood eczema on the parent-child dyad have not been extensively studied in terms of healthrelated quality of life (HRQOL), eczema severity, or mental health status.14

This study was performed to examine associations between mental health issues and disease severity in children or adolescents with AE and their parents using the concise validated 21-item Chinese version of DASS-42 (DASS-21).1 15 16 17

This 2-year cross-sectional survey was conducted between November 2018 and October 2020. Participants with a diagnosis of AE and their parents were recruited at the paediatric dermatology clinic of a university-affiliated hospital in Hong Kong. Eczema was clinically diagnosed in accordance with the United Kingdom Working Party’s Diagnostic Criteria for Atopic Dermatitis.18 Participants and their parents received information about the study including objectives, procedures, voluntary participation, and right of withdrawal. Non-Chinese participants and individuals aged <11 years who were not accompanied by a parent during recruitment were excluded from the study.

The questionnaires were self-administered and supervised by research staff. Parents would help complete the Children’s Dermatology Life Quality Index (CDLQI), Infants’ Dermatitis Quality of Life Index (IDQOL), Nottingham Eczema Severity Score (NESS), and Patient-Oriented Eczema Measure (POEM) for their younger children. The DASS-21 was individually administered to all parents and to children aged >11 years.

The validated Chinese version of the three-item NESS, completed by children aged >11 years or the parents of children aged ≤11 years, was used to determine eczema severity in participating children.19 20 The presence of eczema and number of nights affected by skin itchiness each week in the past 12 months were rated from 1 to 5. A higher score indicated greater eczema severity. Additionally, areas of skin with eczematous lesions (eg, rash, lichenified skin, and/or bleeding) were recorded. Scores on the NESS were categorised as mild (3 to 8), moderate (9 to 11), and severe (12 to 15) eczema. Subjective measurements were determined using the validated seven-item Chinese translation of the POEM, which was also completed by children aged >11 years or the parents of children aged ≤11 years.21 22 Each item was scored from 0 to 4, with a maximum aggregate score of 28. A higher score indicated greater eczema severity in the past week (ranges of 0-2, 3-7, 8-16, 17-24, and 25-28 correspond to clear, mild, moderate, severe, and very severe levels of eczema, respectively).

Health-related QOL was evaluated using the Chinese version of the 10-item CDLQI23 and the 10-item IDQOL.24 The CDLQI was completed by children aged ≥4 years with guidance from parents, whereas the IDQOL was completed by parents of children aged <4 years. Each item on the two scales was scored from 0 to 3, with a maximum aggregate score of 30. A higher score indicated greater eczema-related HRQOL impairment in the past week (CDLQI ranges of 0-1, 2-6, 7-12, 13-18, and 19-30 correspond to no, small, moderate, very large, and extremely large effects on HRQOL, respectively; these ranges for IDQOL are 0-1, 2-5, 6-10, 11-20, and 21-30, respectively).

Mental health was assessed by measuring depression, anxiety, and stress in children with eczema and their parents using the validated Chinese version of the DASS-21. This scale has been used to examine symptoms of depression, anxiety, and stress among individuals with dermatitis or eczema.1 17 The DASS-21 was individually administered to all parents and to children aged >11 years. The DASS-21 composite score can be divided into the DASS Depression, DASS Anxiety, and DASS Stress domains. The total score for each domain ranges from 0 to 42.25 A higher score indicates greater emotional distress in that domain (DASS Depression ranges of 0-9, 10-13, 14-20, 21-27, and ≥28 correspond to normal, mild, moderate, severe, and extremely severe levels, respectively; these ranges for DASS Anxiety are 0-7, 8-9, 10-14, 15-19, and ≥20, whereas they are 0-14, 15-18, 19-25, 26-33, and ≥34 for DASS Stress).

Clinical data were de-identified and analysed using SPSS (Windows version 25.0; IBM Corp, Armonk [NY], United States). Frequency distributions were used to describe the demographic and clinical characteristics of participants. Continuous variables with normal distributions were expressed as means±standard deviations (corrected to 1 decimal place). Nominal and ordinal variables were expressed in numbers with percentage (corrected to 1 decimal place). Independent samples t tests were used to explore sex differences regarding age, education level, disease severity, QOL, and emotional distress in parents and children. Pearson correlation analysis was utilised to examine associations among mental health, eczema severity, and HRQOL in parents and children of both sexes. Multiple linear regression was performed to adjust for variations in parental and child DASS scores and HRQOL according to demographic and clinical variables. P values <0.05 were considered statistically significant.

Among 380 parents (mean age=41.13±6.52 years), 49 were fathers (mean age=42.50±7.49 years) and 331 were mothers (mean age=40.95±6.36 years). Parents’ education levels were primary to secondary (n=124, 13 males), post-secondary (n=26, 1 male), and undergraduate or above (n=81, 14 males) [Table 1]. Parents reported moderate to extremely severe depression (n=58, 2 males), moderate to extremely severe anxiety (n=101, 5 males), and moderate to extremely severe stress (n=90, 6 males). Parents with a higher education level had lower levels of depression, anxiety, and stress (Fig 1). Compared with fathers, mothers generally had higher overall DASS-21 depression, anxiety, and stress scores (P<0.001-0.005) [Table 1].

Among 432 children (mean age=9.61±5.41 years), 218 were boys (mean age=9.15±5.44 years) and 214 were girls (mean age=10.06±5.35 years). Most children had moderate to severe/very severe disease according to the POEM (n=290) and NESS (n=291). Over half of the children displayed a moderate to extremely large impact on QOL in the CDLQI (n=171, 50.4%) and IDQOL (n=56, 62.9%). Small numbers of children had moderate to extremely severe depression (n=36), anxiety (n=43), and stress (n=30). There were no significant sex differences in disease severity, HRQOL, or emotional distress in the DASS (Table 2).

Disease severity in terms of NESS, POEM, and HRQOL (ie, CDLQI and IDQOL) was generally worse among infants than among older children (Fig 2). Thus, eczema severity and QOL generally appeared to improve with age. Correlation analysis demonstrated that depression, anxiety, and stress levels were significantly associated with NESS, POEM, and CDLQI, regardless of sex (Table 3).

Correlation analysis revealed that eczema severity (NESS and POEM) and HRQOL (CDLQI) were associated with depression, anxiety, and stress levels (DASS-21) among children and parents, regardless of sex (Table 3 and Fig 3). Moreover, depression, anxiety, and stress levels in mothers were significantly correlated with NESS, POEM, IDQOL, and CDLQI. Paternal anxiety and stress levels were correlated with NESS, POEM, and CDLQI (P<0.001 to 0.034). However, paternal depression was only correlated with POEM (P=0.014) [Table 3].

Maternal depression showed a marginal association with higher anxiety levels in boys with eczema (n=42, r=0.311; P=0.045) [Table 3]. However, considering the small number of pairs, no clinical or statistical inferences should be made regarding sex differences in mental health among children and parents. Additionally, there were no statistically significant associations between the mental health of children and parents concerning depression, anxiety, and stress levels in the DASS-21 (Table 3). Regression analysis showed that the child’s HRQOL and parental age mostly explained variation in parental anxiety and stress, whereas parental education level explained variation in parental depression (Table 4). Younger parents had higher risk of developing more anxiety and stress compared with the older parents. Depression level of parents with primary to secondary education was 58% higher than their counterparts with post-secondary education or above. Conversely, the child’s eczema severity and HRQOL mostly explained the child’s emotional distress. Eczema severity and parental emotional distress significantly affected HRQOL in children of all ages (Table 4).

There was no psychological or physiological discomfort resulted from administration of the surveys.

Psychological symptoms of depression, anxiety, and stress were prevalent among children with AE and their parents. Our findings indicate associations between the mental health of children and parents and the eczema severity in those children. Increased eczema severity in children and adolescents led to greater emotional distress in parents and children, regardless of sex. Similarly, psychological symptoms in children and their parents were negatively correlated with the child’s eczema severity (NESS and POEM) and HRQOL impairment (CDLQI or IDQOL), regardless of sex. These strong correlations suggest that psychological symptoms, eczema severity, and impact on QOL have mutually detrimental effects. The DASS depression, anxiety, and stress scores were generally higher among mothers than among fathers, suggesting that mothers (the primary caregivers for children with eczema) were more strongly affected. The present study showed that eczema severity can adversely affect emotions and QOL among parents and children, highlighting the need for further exploration of biopsychosocial interactions among children and adolescents with eczema. Children with severe disease reportedly have more problems with depression and internalising behaviour.26 Behavioural issues can lead to adverse social interactions with peers, further reducing self-esteem and HRQOL. Therefore, interactions among parental perception of the child’s disease severity, the child’s treatment adherence, the child’s social influence by peers, and the child’s school environment should be considered when clinicians make comprehensive decisions about holistic treatments.

Our results using the DASS-21 are consistent with findings in previous studies1 27 that used the more comprehensive DASS-42. As in previous studies,1 27 we found that caregivers were especially likely to experience anxiety related to care provision in the home.28 29 In the present study, maternal depression was associated with a higher anxiety level, particularly in relation to boys with eczema. Accordingly, the Harmonising Outcome Measures for Eczema initiative recommends documentation of disease severity and QOL impairment in eczema cases.25 30 However, there have been few international initiatives and clinical trials regarding the psychological symptoms of caregivers and patients, particularly in the context of childhood eczema. Therefore, we suggest that clinicians should consider these important measurable domains in terms of therapeutic interventions and psychological support. Childhood eczema treatments mainly focus on pharmacological control of physical symptoms, but they often completely neglect the psychological symptoms of affected children and their parents. A more holistic treatment approach is needed for this potentially devastating common childhood disorder. Given the increasing numbers of proposed assessment tools, we advocate a holistic and comprehensive approach for eczema management that considers children and their families. This treatment tool should use a composite score to continuously evaluate disease severity (in objective and subjective manners), QOL impairment, psychological symptoms, and miscellaneous disease surrogates in affected children and their parents.1 16 21 26

A strength of this study was that compared with the DASS-42, the DASS-21 demonstrated better performance with 50% fewer questions and a shorter completion time. Findings from the DASS-21, but not the DASS-42, were correlated with disease severity as measured by the NESS and POEM.1 These discrepancies could have arisen because the sample size in the present study (using the DASS-21) was threefold greater than the sample size in the previous DASS-42 study.1 In the present study, the DASS-21, especially in child and mother, was moderately to strongly correlated with the CDLQI, IDQOL, NESS, and POEM. Thus, the DASS-21 can effectively represent the degree of emotional distress among parents and children or adolescents with eczema. This questionnaire is available in different languages, potentially allowing it to be used for assessment of patients with other ethnicities. To our knowledge, this is the first study to use the DASS-21 to assess the mental health of parents and children with eczema in a paediatric setting. This study revealed the presence of childhood eczema-related depression, anxiety, and stress in affected children and their parents.

This study had a few limitations including its relatively small sample size, especially concerning father-child pairs. A greater proportion of mothers participated in this study, which is expected because mothers are the main caregivers for children with eczema; they typically accompany their children during medical consultations. Considering that paediatric dermatological clinics also cater adolescent patients aged ≥16 years, a few participants aged 16 to 19 years completed the CDLQI on their HRQOL; although these participants exceeded the suggested age range of ≤16 years, the overall results were not affected.

Another limitation is that the number of recruited mothers, who are normally regarded as the main child caregiver, much outweighs that of recruited fathers. In addition, compared with fathers, mothers may know their child’s health more and get anxious or depressed as the eczema severity of their child escalates over time. Thus, the difference of the role in childbearing, sample size and the understanding of child’s health may affect the findings in parental-child correlations. It should be cautious when the results regarding parental-child correlations are studied and presented. The CDLQI (n=339) is a questionnaire for children, and the IDQOL (n=89) is for infants. The different numbers of participants who completed each of these questionnaires is consistent with the CDLQI coverage of a broader age range, whereas the IDQOL is only suitable for children aged <4 years. Although maternal depression was correlated with boys with anxiety, it is important to note that statistical significance should not be used to infer that there is a sex difference between parent and child groups in terms of mental health; such an inference would constitute overgeneralisation.

Children with eczema and their parents demonstrated mental health impairment, which was correlated with disease severity. Eczema-induced anxiety, stress, and other mental health issues in affected children and their parents should be considered by healthcare professionals during comprehensive assessments for the treatment of eczema. In addition to primary eczema, possible secondary psychiatric symptoms should be monitored in children with moderate to severe eczema and their parents. Childhood eczema severity and the mental health of affected children and their parents should be simultaneously evaluated to prevent and manage secondary psychological problems.

Concept or design: KL Hon.
Acquisition of data: PH Lam.
Analysis or interpretation of data: PH Lam, P Ip.
Drafting of the manuscript: PH Lam, KL Hon.
Critical revision of the manuscript for important intellectual content: KL Hon, S Loo, MJ Koh, CHY Chan, CK Li, P Ip.

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

The authors thank all children and parents who participated in this research.

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 Joint Chinese University of Hong Kong–New Territories East Cluster Clinical Research Ethics Committee, Hong Kong (Ref No.: CRE.2018.401). Written informed consent was obtained from participants and parents prior to the research.

1. Hon KL, Pong NH, Poon TC, et al. Quality of life and psychosocial issues are important outcome measures in eczema treatment. J Dermatolog Treat 2015;26:83-9. Crossref

2. Leung AK, Hon KL, Robson WL. Atopic dermatitis. Adv Pediatr 2007;54:241-73. Crossref

3. Leung TN, Hon KL. Eczema therapeutics in children: what do the clinical trials say? Hong Kong Med J 2015;21:251-60. Crossref

4. Hon KL, Lam PH, Ng WG, et al. Age, sex, and disease status as determinants of skin hydration and transepidermal water loss among children with and without eczema. Hong Kong Med J 2020;26:19-26. Crossref

5. Hon KL, Wong KY, Leung TF, Chow CM, Ng PC. Comparison of skin hydration evaluation sites and correlations among skin hydration, transepidermal water loss, SCORAD index, Nottingham Eczema Severity Score, and quality of life in patients with atopic dermatitis. Am J Clin Dermatol 2008;9:45-50. Crossref

6. Holm EA, Wulf HC, Stegmann H, Jemec GB. Life quality assessment among patients with atopic eczema. Br J Dermatol 2006;154:719-25. Crossref

7. Slattery MJ, Essex MJ, Paletz EM, et al. Depression, anxiety, and dermatologic quality of life in adolescents with atopic dermatitis. J Allergy Clin Immunol 2011;128:668-71. Crossref

8. Magin PJ, Pond CD, Smith WT, Watson AB, Goode SM. Correlation and agreement of self-assessed and objective skin disease severity in a cross-sectional study of patients with acne, psoriasis, and atopic eczema. Int J Dermatol 2011;50:1486-90. Crossref

9. Kuniyoshi Y, Kikuya M, Miyashita M, et al. Severity of eczema and mental health problems in Japanese schoolchildren: the ToMMo Child Health Study. Allergol Int 2018;67:481-6. Crossref

10. Wan J, Takeshita J, Shin DB, Gelfand JM. Mental health impairment among children with atopic dermatitis: a United States population-based cross-sectional study of the 2013-2017 National Health Interview Survey. J Am Acad Dermatol 2020;82:1368-75. Crossref

11. Fishbein AB, Cheng BT, Tilley CC, et al. Sleep disturbance in school-aged children with atopic dermatitis: prevalence and severity in a cross-sectional sample. J Allergy Clin Immunol Pract 2021;9:3120-9.e3. Crossref

12. Schmitt J, Apfelbacher C, Chen CM, et al. Infant-onset eczema in relation to mental health problems at age 10 years: results from a prospective birth cohort study (German Infant Nutrition Intervention plus). J Allergy Clin Immunol 2010;125:404-10. Crossref

13. Walker C, Papadopoulos L, Hussein M. Paediatric eczema and psychosocial morbidity: how does eczema interact with parents’ illness beliefs? J Eur Acad Dermatol Venereol 2007;21:63-7. Crossref

14. Ali F, Vyas J, Finlay AY. Counting the burden: atopic dermatitis and health-related quality of life. Acta Derm Venereol 2020;100:adv00161. Crossref

15. Wong KC. Psychometric investigation into the construct of neurasthenia and its related conditions: a comparative study on Chinese in Hong Kong and Mainland China [dissertation]. Hong Kong: The Chinese University of Hong Kong; 2009.

16. Lam PH, Hon KL, Leung KK, Leong KF, Li CK, Leung TF. Self-perceived disease control in childhood eczema. J Dermatolog Treat 2022;33:1459-64. Crossref

17. Gong X, Xie XY, Xu R, Luo YJ. Psychometric properties of the Chinese versions of DASS-21 in Chinese college students [in Chinese]. Chinese J Clin Psychol 2010;18:443-6.

18. Williams HC, Burney PG, Pembroke AC, Hay RJ. The U.K. Working Party’s Diagnostic Criteria for Atopic Dermatitis. III. Independent hospital validation. Br J Dermatol 1994;131:406-16. Crossref

19. Emerson RM, Charman CR, Williams HC. The Nottingham Eczema Severity Score: preliminary refinement of the Rajka and Langeland grading. Br J Dermatol 2000;142:288-97. Crossref

20. Hon KL, Ma KC, Wong E, Leung TF, Wong Y, Fok TF. Validation of a self-administered questionnaire in Chinese in the assessment of eczema severity. Pediatr Dermatol 2003;20:465-9. Crossref

21. Hon KL, Kung JS, Tsang KY, Yu JW, Cheng NS, Leung TF. Do we need another symptom score for childhood eczema? J Dermatolog Treat 2018;29:510-4. Crossref

22. Gaunt DM, Metcalfe C, Ridd M. The Patient-Oriented Eczema Measure in young children: responsiveness and minimal clinically important difference. Allergy 2016;71:1620-5. Crossref

23. Salek MS, Jung S, Brincat-Ruffini LA, et al. Clinical experience and psychometric properties of the Children’s Dermatology Life Quality Index (CDLQI), 1995-2012. Br J Dermatol 2013;169:734-59. Crossref

24. Lewis-Jones MS, Finlay AY, Dykes PJ. The Infants’ Dermatitis Quality of Life Index. Br J Dermatol 2001;144:104-10. Crossref

25. Gerbens LA, Prinsen CA, Chalmers JR, et al. Evaluation of the measurement properties of symptom measurement instruments for atopic eczema: a systematic review. Allergy 2017;72:146-63. Crossref

26. Hon KL, Kam WY, Lam MC, Leung TF, Ng PC. CDLQI, SCORAD and NESS: are they correlated? Qual Life Res 2006;15:1551-8. Crossref

27. Duran S, Atar E. Determination of depression, anxiety and stress (DAS) levels in patients with atopic dermatitis: a case-control study. Psychol Health Med 2020;25:1153-63. Crossref

28. Shelley AJ, McDonald KA, McEvoy A, et al. Usability, satisfaction, and usefulness of an illustrated eczema action plan. J Cutan Med Surg 2018;22:577-82. Crossref

29. Rork JF, Sheehan WJ, Gaffin JM, et al. Parental response to written eczema action plans in children with eczema. Arch Dermatol 2012;148:391-2. Crossref

30. Spuls PI, Gerbens LA, Simpson E, et al. Patient-Oriented Eczema Measure (POEM), a core instrument to measure symptoms in clinical trials: a Harmonising Outcome Measures for Eczema (HOME) statement. Br J Dermatol 2017;176:979-84. Crossref

Despite high-quality care, many patients admitted to the intensive care unit (ICU) do not survive; therefore, management of the dying process is a required skill among modern healthcare professionals.1 Life-sustaining technology has advanced sufficiently that it is possible to maintain vital organ function despite the knowledge that the patient’s return to health and an acceptable quality of life is no longer feasible. In these situations, a decision to limit life-sustaining treatment (LST) has become a common clinical practice in most countries worldwide.2 3 4 5 6 In recent decades, attempts to define desirable principles for end-of-life care according to a global professional consensus have achieved considerable success.3 Nevertheless, decision-making processes for death and dying are likely to be heavily influenced by regional and cultural norms and expectations; thus, it is reasonable to expect different medical practices related to end-of-life decisions. Several local and international surveys of healthcare professionals have revealed regional differences in attitudes towards end-of-life ethical concerns, as well as substantial differences in clinical practices.7 8 9 10 11 Limited prospective observational data from international studies support the existence of regional variability in end-of-life practices.5 12 13

Hong Kong is a special administrative region of China with an overwhelmingly Chinese population; nevertheless, it maintains an independent fiscal budget and healthcare system. The Hong Kong Hospital Authority, funded by the Hong Kong SAR Government, provides >90% of hospital-based services available for the local population; although nearly all healthcare workers in the public health services exhibit Chinese ethnicity, health services are based on Western medical conventions.14 Hong Kong is considered a high-income region, and recently published patient outcomes data indicate that the Hong Kong Hospital Authority provides high-quality intensive care services.15 The juxtaposition of a Western medical system and a culturally Chinese population creates a situation where Western medical practices (driven by Western cultural and ethical values) may conflict with Chinese cultural values, particularly at the end of life when deep-rooted cultural beliefs may become more relevant. A small number of studies have explored end-of-life care practices in Hong Kong ICUs; these include a survey of ICU physicians’ ethical attitudes concerning end-of-life care8 and a prospective observational study regarding end-of-life practices at a single tertiary university hospital.16 No observational territory-wide data have been published thus far. Additionally, end-of-life practices in Europe have substantially changed in recent decades17; similar changes may have occurred in Hong Kong, although previous comparative data are sparse.16 Multiple Hong Kong ICUs participated in the recent worldwide Ethicus-2 study,13 18 with the understanding that the Hong Kong data would be accessible for secondary analysis. The aim of this study was to provide a detailed description of current end-of-life care practices in Hong Kong.

This study constituted a secondary analysis of the Ethicus-2 database, focusing on the Hong Kong data. The Ethicus-2 study was a prospective, multicentre, global observational study of end-of-life practices in 199 ICUs across 36 countries.13 17 All 15 adult ICUs in publicly funded hospitals in Hong Kong were invited to participate by the Hong Kong study coordinator, representing the Hong Kong Society of Critical Care Medicine. Eight ICUs in Hong Kong participated.

Consecutive adult patients admitted to the ICU over an individual ICU-selected 6-month period between 1 September 2015 and 30 September 2016 with LST limitation or death were included. Follow-up continued until death or 2 months from the initial decision to limit LST. End-of-life categories included withholding LST, withdrawing LST, active shortening of the dying process, failed cardiopulmonary resuscitation (CPR), and brain death. These categories were mutually exclusive; if more than one limitation was triggered in a particular case, the most stringent limitation was chosen (ie, active shortening of the dying process was considered more stringent than LST withdrawal, followed by LST withholding).

Data were collected by the senior physician, or a representative, responsible for making end-of-life decisions. De-identified patient data were entered into a secure online database. Collected data included age; sex; religion; end-of-life category; admission date, time, and diagnoses; chronic disorders; use of ventilation and vasopressors, sedatives, or analgesics; date and time of hospital and ICU admission; and date and time of death or discharge from the ICU or hospital. End-of-life process data collected included type, date, and time of LST; presence of information about patient wishes; discussions with the patient or their family; degree of concurrence between the decision and patient/family wishes; and reasons for treatment decisions.

Data quality was monitored by concurrent audit and feedback, with a quality review involving 5% of all patients.17 Categorical variables were reported as numbers and percentages within end-of-life groups. After normality assessment using the Shapiro–Wilk test, continuous variables were reported as means (standard deviations) or medians (interquartile ranges [IQRs]), as appropriate. Differences among LST withholding, LST withdrawal, and no LST limitation groups were compared using analysis of variance, the Kruskal–Wallis H test, or the Chi squared test, as appropriate. Subsequent pairwise group comparisons were performed with Bonferroni correction for multiple tests. All analyses were performed using SPSS software (Windows version 27.0; IBM Corp, Armonk [NY], United States).

This prospective observational study has been reported in accordance with the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) checklist for observational studies.

The eight participating ICUs were distributed across Hong Kong; at least two ICUs represented each of the New Territories, Kowloon, and Hong Kong Island. Two ICUs were located in academic university hospitals (comprising 20 and 25 acute ICU beds, respectively), and the remainder were located in medium-to-large regional hospitals (ranging from 12 to 22 acute ICU beds per unit).

Among the 4922 consecutive patients screened during the study period, 548 (11.1%) patients with LST limitation (withholding or withdrawal) or death (failed CPR or brain death) were included in the study. Life-sustaining treatment limitation occurred in 455 (83.0%) patients, including 353 (77.6%) with decisions to withhold LST and 102 (22.4%) with decisions to withdraw LST. Of the 93 patients who died without LST limitation, 80 (86.0%) had failed CPR, and 13 (14.0%) experienced brain death (Fig 1). No patients underwent shortening of the dying process.

Patient characteristics are summarised in Table 1; knowledge of patient and family/surrogate wishes, as well as the timing of end-of-life processes, are described in Table 2. Patients without LST limitation had a shorter duration of ICU stay (median: 3 days, IQR=1-6) compared with patients who had decisions to withhold (median: 4 days, IQR=2-13) or withdraw (median: 6 days, IQR=3-11) [P<0.001].

The prevalences of treatments withheld or withdrawn at the initial and final decisions to limit LST are shown in Figure 2. Higher percentages of patients had endotracheal tube (P=0.009), renal replacement therapy (P<0.001), and sedation/analgesia (P=0.002) withheld at the final decision, compared with the initial decision. Similarly, higher percentages of patients had endotracheal tube, mechanical ventilation, vasopressor, and renal replacement therapy withdrawn at the final decision (all P<0.001), compared with the initial decision.

Information about decision-making practices for patients with LST limitation is provided in Table 3. In the majority of cases, the ICU physician was involved in key aspects of end-of-life decision-making and implementation. The responsible ICU physicians’ explanations of the reasons and considerations for supporting end-of-life decisions are provided in Table 4. The primary clinical reason for limiting LST was unresponsiveness to maximal therapy; the patient’s best interest, perceived good medical practice, and autonomy were key decision-making considerations.

This is the first large, multicentre, prospective, observational study of end-of-life care practices in Hong Kong ICUs. Our main findings were that LST limitation preceded >80% of patient deaths, and that death occurred in the vast majority of patients with LST limitation; only 4% of patients with LST limitation were alive at 2 months. Only 15% of ICU patients died after failed CPR (ie, without any LST limitations). Advance directives were rarely available, and no cases of active shortening of the dying process (euthanasia) were reported. Life-sustaining treatment limitation occurred in the majority (83.0%) of patients, predominantly via withholding (77.6%); withdrawal was less common (22.4%) [Fig 1]. High rates of LST limitation, such as those observed in this study, are generally presumed to reflect good end-of-life practices and have been associated with the presence of written end-of-life guidelines,19 such as those provided by the Hong Kong Hospital Authority.20

Regarding treatments that were withdrawn or withheld, the withholding of CPR universally accompanied all limitation decisions. Nutrition, hydration, and sedation were rarely withheld or withdrawn at any time, consistent with guidance from professional bodies in Hong Kong that additional safeguards are necessary when considering these actions.20 At the time of the initial limitation decision, there was relatively frequent withholding of vasopressors and renal replacement therapy; withholding or withdrawal of endotracheal tubes was less common. Although the patterns of LST limitation were similar between the initial and final decisions, such that withholding remained more prevalent than withdrawal, a substantial increase was observed in the prevalence of LST withdrawal at the time of the final decision. This finding may reflect the common Chinese cultural perspective that LST withholding and withdrawal are not ethically equivalent, with a documented preference for withholding over withdrawal as an end-of-life care strategy.8 11 The increase in withdrawal prevalence at the time of the final decision across key treatment categories (eg, vasopressors, mechanical ventilation, and renal replacement therapy) suggests that, with increasing prognostic certainty and clear progression towards death, LST withdrawal becomes more acceptable. There also appeared to be a greater reluctance to adopt withdrawal strategies early in the ICU stay, evidenced by the longer interval between ICU admission and initial limitation, if the initial limitation was withdrawal. This tendency may also reflect the need for greater prognostic certainty prior to the implementation of a withdrawal strategy. Comparisons with international data indicate that although the high rate of LST limitation prior to death is similar to practices in other countries, the early and more frequent use of withholding (rather than withdrawal) remains distinct from practices reported in North America, North and Central Europe, and Asia.13

Comparative historical data from Hong Kong are limited. A single-centre observational study conducted between 1997 and 1999 showed that LST limitation occurred in 59% of patients,16 although its LST limitation categories are not fully aligned with those of the current study. Notably, the mean interval between ICU admission and LST limitation in this previous study was nearly 8 days,16 whereas the median interval in the current study was 1.8 days (IQR=0.5-7); this difference suggests that recognition of the need for LST limitation is occurring much earlier in Hong Kong, consistent with a pattern observed in Europe during the same period.17 21 When LST limitation is indicated, earlier intervention leads to a shorter duration of patient discomfort; the observed reduction in time to limitation may represent a meaningful practice improvement over time.

Despite similar rates of LST withholding/withdrawal, the low rate of survival after LST limitation in Hong Kong (3%-4%)—comparable to the findings in a previous pan-European study12—contrasts with current European ICU outcomes, where the combined survival rate after LST withdrawal or withholding was 20%.17 This difference may possibly be attributed to implementing LST in patients at the very end-of-life when prognostic certainty is greater. The earlier implementation of end-of-life interventions may represent an area for further exploration to improve end-of-life ICU practices and minimise suffering.

Key practices in end-of-life decision-making included the initiation of discussions to limit LST by ICU physicians in the vast majority of cases; when such discussions began, ICU physicians were always involved in end-of-life decision-making processes. Notably, shared decision-making between ICU physicians and families was the predominant model reported. These findings align with the best practices described in recent international expert consensus documents.1 3 Despite frequent use of the shared decision-making model, direct or indirect knowledge of the patient’s wishes regarding LST was available for fewer than half of patients (40.3%); in the vast majority of cases (94.6%), this information was transmitted by relatives rather than by the patient themselves. Only 33 (6.0%) patients had decision-making capacity during the decision-making process, and only two (0.4%) patients had advance directives (Table 2). These results highlight the need to encourage patients to discuss their wishes regarding future end-of-life care preferences with relatives, or communicate such wishes through the use of advance directives, ensuring that patients receive the preferred level of care at this critical time. Nevertheless, levels of agreement among all parties regarding end-of-life decisions were high, and delays in decision-making due to disagreement were uncommon.

Advance directives in Hong Kong ICUs were rarely available, possibly due to selection bias; individuals with advanced disease and a greater likelihood of advance directives may have lower ICU admission priority. However, the current rate of advance directive use in North American ICUs at the end of life is nearly 50%.18 A relatively recent population-based study demonstrated very low public awareness of advance directives in Hong Kong, such that 86% of participants reported no previous knowledge of the advance directive concept.22 However, once informed of this concept, the majority of participants indicated a willingness to consider using such directives. The legislative process to formalise advance directive use in Hong Kong has substantially progressed, and there is a recognised need for public education and healthcare professional–specific guidance to promote the use of these directives.23 24

In the present study, the most common diagnostic categories at ICU admission were respiratory (43.4%) and sepsis-related (38.0%) [Table 1], similar to reported findings in most other regions worldwide.18 There were no substantial age or sex differences regarding LST limitation, but there were distinct differences in ICU admission diagnoses, such that limitation was less likely in patients with cardiovascular conditions and more likely in patients with sepsis or gastrointestinal disease. The vast majority of patients exhibited at least one co-morbidity, again similar to recently reported findings in other regions.18 Intriguingly, no patients with cancer were among those who died without LST limitation.

The primary clinical reasons for initiating LST limitation included unresponsiveness to maximal therapy, multiorgan failure, and neurologic failure; in few cases, the limitation arose from a family request or mainly in relation to quality of life (Table 4). Overwhelmingly, the primary consideration for decision-making was the patient’s best interest, followed by the principle of good medical practice, defined as the recognition that continued maximal therapy would not be beneficial for the patient (Table 4). These observations closely match the responses recently provided by a group of international experts who were asked to rank the triggers they would likely use in clinical practice to initiate discussions about LST limitation.1

Two questions related to decision-making and patient treatment wishes revealed an interesting observation. Across all end-of-life categories, approximately 70% of physicians in charge of end-of-life decision-making reported that if the patient’s wishes were known, they were followed. In contrast, when a surrogate’s treatment wishes were known, they were followed in nearly every case (Table 2). These responses indicate that the family’s treatment preferences are respected more frequently compared with known patient preferences, in contrast to guidelines from the Medical Council of Hong Kong25 and the Hospital Authority.20 Both guidelines clearly state that treatment preferences should be sought via communication with patients and family when possible, and a consensus should be reached; however, when conflicting views cannot be reconciled, the patient’s treatment preferences should supersede the family’s preferences.20 25 It is possible that physicians prioritised the family’s preferences because few patients were capable of direct communication; there was low certainty regarding perceived patient wishes when communicated through third parties. Nevertheless, this finding warrants further investigation and reflection among Hong Kong ICU healthcare professionals.

Most communication related to end-of-life decision-making occurred between ICU physicians and family/surrogates; nurses, primary physicians, and consulting physicians were involved in fewer than half of the reported cases. It has been suggested that this relatively low percentage of nurse involvement is an underestimate because most data were reported by physicians who may be unaware of nurse involvement.26 Decision agreement between healthcare staff and family members, as well as among family members, was reportedly very high (>95%) [Table 3]. Disagreements between family and staff were rare, as were delays in implementing end-of-life care because of disagreement, indicating a high level of acceptance of the decision-making process by the public and healthcare professionals in Hong Kong.

This study’s strengths included its involvement of a large number of patients over a 6-month period, provision of detailed follow-up data for up to 2 months, prospective design, and representation of most public ICUs in Hong Kong. Moreover, the data were provided by the physician in charge of end-of-life decision-making, with support from clear definitions and uniform collection across ICUs; they were also subjected to external quality control measures, minimising measurement bias. The main limitations were the lack of random ICU allocation and inclusion of consenting ICUs only, which may have introduced selection bias.

Data from the majority of Hong Kong ICUs, spanning the entire territory and representing both academic and non-academic ICUs, revealed that LST limitation occurs in most patients prior to death in ICU. Practices generally align with recommendations from local professional bodies and key international consensus documents. Although decision-making is usually initiated by ICU physicians, shared decision-making between medical staff and family/surrogates is the predominant model. Because a loss of decision-making capacity is common in the ICU, patients should be encouraged to communicate their wishes regarding end-of-life care through dialogue with relatives or more formal methods. Certain practices and outcomes observed in Hong Kong are more similar to those reported in North America and Europe than to patterns in other parts of Asia.

Concept or design: CL Sprung, A Avidan, GM Joynt.
Acquisition of data: GM Joynt, SKH Ling, LL Chang, PNW Tsai, GKF Au, DHK So, FL Chow, PKN Lam.
Analysis or interpretation of data: A Lee, GM Joynt.
Drafting of the manuscript: GM Joynt.
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.

All authors have disclosed no conflicts of interest.

No funding was obtained to conduct this research from any funding agency in the public, commercial, or not-for-profit sectors. The Walter F and Alice Gorham Foundation funded the international co-ordination of the Ethicus-2 study. The Foundation had no part in the design and conduct of the research; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

This research was approved by the relevant Research Ethics Committee for each of the participating centres, including:
(1) Tuen Mun Hospital—The New Territories West Cluster Research Ethics Committee of Hospital Authority, Hong Kong (Ref No.: NTWC/CREC/15078);
(2) Prince of Wales Hospital and North District Hospital—The Joint Chinese University of Hong Kong—New Territories East Cluster Clinical Research Ethics Committee, Hong Kong (Ref No.: 2015.080);
(3) Caritas Medical Centre—The Kowloon West Cluster Research Ethics Committee of the Hospital Authority, Hong Kong [Ref No.: KW/EX-15-103(88-02)];
(4) Kwong Wah Hospital—The Kowloon West Cluster Research Ethics Committee of the Hospital Authority, Hong Kong [Ref No.: KW/EX-15-105(88-04)];
(5) Pamela Youde Nethersole Eastern Hospital—The Hong Kong East Cluster Research Ethics Committee of the Hospital Authority, Hong Kong (Ref No.: HKEC-2015-028);
(6) Princess Margaret Hospital—The Kowloon West Cluster Research Ethics Committee of the Hospital Authority, Hong Kong [Ref No.: KW/EX-15-104(88-03)]; and
(7) Queen Mary Hospital—Institutional Review Board of The University of Hong Kong / Hospital Authority Hong Kong West Cluster, Hong Kong (Ref No.: UW 15-361).

The requirement for informed patient consent was waived by the relevant Clinical Research Ethics Committees as the study was observational only, where all collected data were anonymised at source and only de-identified data were passed on to the co-ordinating centre for analysis, and risk to participants was minimal.

Gary KF Au, Department of Intensive Care, Kwong Wah Hospital, Hong Kong SAR, China
Alexander Avidan, Department of Anesthesiology, Critical Care and Pain Medicine, Hadassah Medical Organization and Faculty of Medicine, Hebrew University of Jerusalem, Jerusalem, Israel
KC Chan, Department of Intensive Care, Tuen Mun Hospital, Hong Kong SAR, China
WM Chan, Adult Intensive Care Unit, Queen Mary Hospital, Hong Kong SAR, China
LL Chang, Department of Intensive Care, Pamela Youde Nethersole Eastern Hospital, Hong Kong SAR, China
FL Chow, Department of Intensive Care, Caritas Medical Centre, Hong Kong SAR, China
Gavin Matthew Joynt, Department of Anaesthesia and Intensive Care, The Chinese University of Hong Kong, Prince of Wales Hospital, Hong Kong SAR, China
Gladys WM Kwan, Department of Intensive Care, Tuen Mun Hospital, Hong Kong SAR, China
Philip KN Lam, Department of Intensive Care, North District Hospital, Hong Kong SAR, China
Anna Lee, Department of Anaesthesia and Intensive Care, The Chinese University of Hong Kong, Prince of Wales Hospital, Hong Kong SAR, China
E Leung, Adult Intensive Care Unit, Queen Mary Hospital, Hong Kong SAR, China
Steven KH Ling, Department of Intensive Care, Tuen Mun Hospital, Hong Kong SAR, China
CH Ng, Department of Intensive Care, Kwong Wah Hospital, Hong Kong SAR, China
HP Shum, Department of Intensive Care, Pamela Youde Nethersole Eastern Hospital, Hong Kong SAR, China
Dominic HK So, Department of Intensive Care Unit, Princess Margaret Hospital/Yan Chai Hospital, Hong Kong SAR, China
Charles L Sprung, Department of Anesthesiology, Critical Care and Pain Medicine, Hadassah Medical Organization and Faculty of Medicine, Hebrew University of Jerusalem, Jerusalem, Israel
L Sy, Department of Intensive Care, North District Hospital, Hong Kong SAR, China
Polly NW Tsai, Adult Intensive Care Unit, Queen Mary Hospital, Hong Kong SAR, China
HH Tsang, Department of Intensive Care, Kwong Wah Hospital, Hong Kong SAR, China
WT Wong, Department of Anaesthesia and Intensive Care, The Chinese University of Hong Kong, Prince of Wales Hospital, Hong Kong SAR, China

1. Joynt GM, Lipman J, Hartog C, et al. The Durban World Congress Ethics Round Table IV: health care professional end-of-life decision making. J Crit Care 2015;30:224-30. Crossref

2. Vincent JL. Forgoing life support in western European intensive care units: the results of an ethical questionnaire. Crit Care Med 1999;27:1626-33. Crossref

3. Sprung CL, Truog RD, Curtis JR, et al. Seeking worldwide professional consensus on the principles of end-of-life care for the critically ill. The Consensus for Worldwide End-of-Life Practice for Patients in Intensive Care Units (WELPICUS) study. Am J Respir Crit Care Med 2014;190:855-66. Crossref

4. Phua J, Joynt GM, Nishimura M, et al. Withholding and withdrawal of life-sustaining treatments in low-middle-income versus high-income Asian countries and regions. Intensive Care Med 2016;42:1118-27. Crossref

5. Lobo SM, De Simoni FH, Jakob SM, et al. Decision-making on withholding or withdrawing life support in the ICU: a worldwide perspective. Chest 2017;152:321-9. Crossref

6. Wong WT, Phua J, Joynt GM. Worldwide end-of-life practice for patients in ICUs. Curr Opin Anaesthesiol 2018;31:172-8. Crossref

7. Attitudes of critical care medicine professionals concerning forgoing life-sustaining treatments. The Society of Critical Care Medicine Ethics Committee [editorial]. Crit Care Med 1992;20:320-6. Crossref

8. Yap HY, Joynt GM, Gomersall CD. Ethical attitudes of intensive care physicians in Hong Kong: questionnaire survey. Hong Kong Med J 2004;10:244-50.

9. Yaguchi A, Truog RD, Curtis JR, et al. International differences in end-of-life attitudes in the intensive care unit: results of a survey. Arch Intern Med 2005;165:1970-5. Crossref

10. Bito S, Asai A. Attitudes and behaviors of Japanese physicians concerning withholding and withdrawal of life-sustaining treatment for end-of-life patients: results from an Internet survey. BMC Med Ethics 2007;8:7. Crossref

11. Weng L, Joynt GM, Lee A, et al. Attitudes towards ethical problems in critical care medicine: the Chinese perspective. Intensive Care Med 2011;37:655-64. Crossref

12. Sprung CL, Cohen SL, Sjokvist P, et al. End-of-life practices in European intensive care units: the Ethicus study. JAMA 2003;290:790-7. Crossref

13. Avidan A, Sprung CL, Schefold JC, et al. Variations in end-of-life practices in intensive care units worldwide (Ethicus-2): a prospective observational study. Lancet Respir Med 2021;9:1101-10. Crossref

14. Kong X, Yang Y, Gao J, et al. Overview of the health care system in Hong Kong and its referential significance to mainland China. J Chin Med Assoc 2015;78:569-73. Crossref

15. Ling L, Ho CM, Ng PY, et al. Characteristics and outcomes of patients admitted to adult intensive care units in Hong Kong: a population retrospective cohort study from 2008 to 2018. J Intensive Care 2021;9:2. Crossref

16. Buckley TA, Joynt GM, Tan PY, Cheng CA, Yap FH. Limitation of life support: frequency and practice in a Hong Kong intensive care unit. Crit Care Med 2004;32:415-20. Crossref

17. Sprung CL, Ricou B, Hartog CS, et al. Changes in end-of-life practices in European intensive care units from 1999 to 2016. JAMA 2019;322:1692-704. Crossref

18. Feldman C, Sprung CL, Mentzelopoulos SD, et al. Global comparison of communication of end-of-life decisions in the ICU. Chest 2022;162:1074-85. Crossref

19. Mentzelopoulos SD, Chen S, Nates JL, et al. Derivation and performance of an end-of-life practice score aimed at interpreting worldwide treatment-limiting decisions in the critically ill. Crit Care 2022;26:106. Crossref

20. Hospital Authority, Hong Kong SAR Government. Working Group on review of Hospital Authority Clinical Ethics Committee (HKCEC) guidelines related to EOL decisionmaking. HA Guidelines on Life-Sustaining Treatment in the Terminally Ill. 2020. Available from: https://www.ha.org.hk/haho/ho/psrm/LSTEng.pdf. Accessed 4 Jul 2023.

21. Lesieur O, Herbland A, Cabasson S, Hoppe MA, Guillaume F, Leloup M. Changes in limitations of life-sustaining treatments over time in a French intensive care unit: a prospective observational study. J Crit Care 2018;47:21-9. Crossref

22. Chung RY, Wong EL, Kiang N, et al. Knowledge, attitudes, and preferences of advance decisions, end-of-life care, and place of care and death in Hong Kong. A population-based telephone survey of 1067 adults. J Am Med Dir Assoc 2017;18:367.e19-27. Crossref

23. Food and Health Bureau, Hong Kong SAR Government. End-of-Life Care: Legislative Proposals on Advance Directives and Dying in Place Consultation Report. July 2020. Available from: https://www.healthbureau.gov.hk/download/press_and_publications/consultation/190900_eolcare/e_EOL_consultation_report.pdf. Accessed 4 Jul 2023.

24. Fong BY, Yee HH, Ng TK. Advance directives in Hong Kong: moving forward to legislation. Ann Palliat Med 2022;11:2622-30. Crossref

25. Medical Council of Hong Kong. Code of professional conduct: for the guidance of registered medical practitioners. Revised October 2022. Available from: https://www.mchk.org.hk/english/code/files/Code_of_Professional_Conduct_(English_Version)_(Revised_in_October_2022).pdf. Accessed 4 Jul 2023.

26. Benbenishty J, Ganz FD, Anstey MH, et al. Changes in intensive care unit nurse involvement in end of life decision making between 1999 and 2016: descriptive comparative study. Intensive Crit Care Nurs 2022;68:103138. Crossref