Recently, an increased incidence of suicide among students has triggered immense public concern regarding the mental health of adolescents and young adults. In 2014, there were 52 suicides in the age-group 15 to 24 years. The number of suicides in this age-group increased in subsequent years to 68 in 2015 and 75 in 2016.1 A report from the Centre for Health Protection in Hong Kong showed that, in 2012, the prevalences of mild, moderate, and severe depressive symptoms in adolescents and children were 36.4%, 14.7%, and 4.2%, respectively.2 The Population Health Survey 2003/2004 conducted collaboratively by the Department of Health and the Department of Community Medicine of the University of Hong Kong revealed that the median score for the State-Trait Anxiety Inventory was highest in those aged 25 to 34 years, whereas the median score for the Centre for Epidemiologic Studies Depression Scale was highest in those aged 15 to 34 years.3 A web-based survey targeting first-year students receiving tertiary education in Hong Kong concluded that, in 2006, the prevalence of depression was 20.9%, while that of anxiety was 41.2%.4 Another recent study, the Hong Kong Mental Morbidity Survey,5 revealed that the most common mental problem in Hong Kong was mixed anxiety and depressive disorder; moreover, there was a strong association between anxiety symptoms and depressive symptoms.

According to the Diagnostic and Statistical Manual of Mental Disorders, published by the American Psychiatric Association, major depressive disorder is defined as the presence of five or more of the listed symptoms for most of the days during the same 2-week period; at least one of the symptoms must be either depressed mood or loss of interest or pleasure. Notably, these symptoms should reflect a change from previous functioning. Generalised anxiety disorder (GAD) is characterised by excessive worries that cause distress and interfere with psychosocial functioning. These worries frequently happen without precipitants, exhibit longer durations, and are accompanied by three or more of the six listed additional symptoms.

Various factors associated with major depressive disorder and GAD have been identified, including alcohol use, illicit drug use, tobacco use, and level of physical activity.6 However, evidence for an association between academic pressure and these two psychiatric disorders remains unknown among university undergraduate students in Hong Kong. We aimed to provide an update regarding the prevalence of depressive and anxiety symptoms experienced by undergraduate students in Hong Kong, and to identify factors associated with these symptoms.

Full-time undergraduate students of the eight universities in Hong Kong that are funded by the University Grants Committee (UGC) were the target participants in this study. A cross-sectional study design was employed and the target number of students to be enrolled from each university was 150; the total target number of students to be recruited was 1200.

The inclusion criteria for this study were as follows: full-time undergraduate students in the eight UGC-funded universities who were ≥18 years, were able to understand the consent, provide a valid oral consent, comprehend the questionnaire, and had not previously participated in this research. Students who failed to fulfil all inclusion criteria were excluded from the study.

Convenience sampling was employed. Questionnaires were distributed at popular locations within all eight university campuses on school days in September and October 2016. Participants were asked to complete three sets of questionnaires, including the 9-item Patient Health Questionnaire (PHQ-9) for screening of depressive symptoms, the 7-item GAD scale (GAD-7) for screening of anxiety symptoms, and a socio-demographic questionnaire to identify factors associated with depressive and anxiety symptoms.7 8 When completed questionnaires were being returned, respondents were reminded that they would not be able to withdraw from the study after returning the questionnaires, as the questionnaires did not contain any personal identifying information to allow individual retrieval.

The PHQ-9 was adopted to screen for depressive symptoms in this study. Both Chinese and English versions were provided. The PHQ-9 has been shown to be a valid and reliable tool for assessing depressive symptoms in the Hong Kong general population.7 8 Another study showed that the PHQ-9 was a useful tool to detect both major depression and subthreshold depression.9 According to the PHQ-9, a score of <5 was defined as none-minimal depressive symptoms, score of 5 to 9 as mild, score of 10 to 14 as moderate, score of 15 to 19 as moderately severe, and score of ≥20 as severe.

The GAD-7 is a practical self-report anxiety questionnaire that is generally used in out-patient and primary care settings for referral to a psychiatrist to confirm the diagnosis of GAD. Both Chinese and English versions were available. The adoption of GAD-7 in this study was due to its good reliability, as well as its criteria, construct, and factorial and procedural validity.10 11 According to the GAD-7, a score of <5 was defined as none-minimal anxiety symptoms, score of 5 to 9 as mild, score of 10 to 14 as moderate, and score of ≥15 as severe.

Descriptive analysis was performed to summarise the demographics and statuses of depressive and anxiety symptoms of the respondents. A binary logistic regression was conducted to ascertain the effects of various covariates on the odds of exhibiting mild to severe depressive symptoms, on the basis of PHQ-9 results. A PHQ-9 score of <5 was defined as no depressive symptoms, while a score of ≥5 was defined as mild to severe depressive symptoms. Another binary logistic regression was used to ascertain the effects of multiple covariates on the odds that participants would exhibit mild to severe anxiety symptoms, on the basis of GAD-7 results. A GAD-7 score of <5 was defined as no anxiety symptoms, while a score of ≥5 was defined as mild to severe anxiety symptoms. A P value of <0.05 was regarded as a significant difference.

In total, 1480 undergraduate students from the eight UGC-funded universities were approached from 6 September 2016 to 3 October 2016. From these, 1200 completed questionnaires were collected, for a response rate of 81.1%. A total of 81 students did not meet at least one of the inclusion criteria and their responses were subsequently excluded from analysis (Fig). The remaining 1119 students fulfilled all inclusion criteria and their responses were analysed by SPSS (Mac version 24; IBM Corp, Armonk [NY], United States).

The mean (standard deviation) age of the respondents was 19.81 (1.48) years (range, 18-29 years). Among the respondents, 426 (38.1%) were male and 693 (61.9%) were female. Most respondents were Chinese (92.7%). The respondents’ demographics are shown in Table 1.

Among the 1119 valid questionnaires analysed, 767 (68.5%) were found to have mild to severe depressive symptoms (Table 2).

For screening of anxiety symptoms, 18 of the 1119 respondents were excluded because they did not complete the GAD-7 questionnaires; therefore, results are based on the 1101 valid responses. A total of 599 (54.4%) respondents were found to have mild to severe anxiety symptoms. A higher prevalence of anxiety was observed in females than males in all three categories of severity (Table 3).

The binary logistic regression model for the effects of multiple covariates on the odds of having mild to severe depressive symptoms, on the basis of the PHQ-9 results, was statistically significant with χ2=375.006, P<0.001, with 18 degrees of freedom (Table 4). The model explained 43.1% of the variance in depression severity, as shown by Nagelkerke R², and correctly classified 79.6% of the cases. Mild to severe anxiety symptoms, as screened by GAD-7, were significantly associated with mild to severe depressive symptoms (P<0.001). In contrast, several lifestyle and psychosocial variables, including regular exercise (P<0.001), self-confidence (P=0.01), satisfaction with academic performance (P=0.019), and optimism towards the future (P<0.001) were inversely related to mild to severe depressive symptoms.

Another binary logistic regression model for the effects of multiple covariates on the odds of having mild to severe anxiety symptoms, on the basis of the GAD-7 results, was statistically significant with χ²=374.842, P<0.001, with 20 degrees of freedom (Table 5). The model explained 41.3% of the variance in depression severity, as shown by Nagelkerke R², and correctly classified 76.9% of the cases. Mild to severe anxiety symptoms were associated with level of academic difficulty (P=0.028) and mild to severe depressive symptoms, as screened by PHQ-9 (P<0.001). Certain lifestyle and psychosocial variables, including satisfaction with friendship (P=0.001), sleep quality (P=0.003), and self-confidence (P=0.003) were inversely associated with mild to severe anxiety symptoms.

The results revealed that 68.5% of respondents had mild to severe depressive symptoms and 54.4% had mild to severe anxiety symptoms. These rates are higher than and comparable to the results of a similar survey conducted by the University of Hong Kong 10 years ago, respectively.4 This could be attributed to the increasing academic pressure after major reforms of the education system in Hong Kong,12 uncertainties in career prospects due to fluctuations in the socio-political environment, and the more prevalent use of social media.13

This study revealed that mild to severe depressive symptoms, as screened by PHQ-9, were associated with mild to severe anxiety symptoms, as screened by GAD-7. This is consistent with previous studies, including a study in patients with depression showing that 85% of those with major depression were also diagnosed with generalised anxiety.14 Sharing common risk factors and symptoms may explain their co-existence.15 16 Another study has shown that anxiety is more frequently an antecedent of depression, but the reverse relationship was not observed.17 This could be explained by the withdrawal and submissive techniques adopted by individuals with anxiety, in response to social exclusion.18 However, a definite causal relationship remains unclear.

Regular exercise could decrease the occurrence of depression via both physiological and psychological mechanisms. Exercise exerts an excitatory effect on the monoamine and endorphin neurotransmitter systems; notably, monoamines are depleted in patients with depression.19 Psychologically, exercise is reported to improve self-esteem and self-perception through self-actualisation and gaining pleasure from an expanded social circle.20

Psychosocial factors, including higher self-confidence, satisfaction with academic performance, and optimism towards the future are inversely related to depressive symptoms. These could be the presentation or consequences of depression; these factors may have protective effects against developing depression.21 Optimism towards the future, apart from personal factors, is also related to community factors. Risk factors include community disadvantage, safety, and discrimination; an important protective factor is community connectedness. These factors were associated with depressive symptoms.22 This implies that, in addition to promoting personal mental health, community-level risk and protective factors that affect individuals’ optimism towards future are sufficiently important to receive broader attention.

For the analysis of anxiety, the level of academic difficulty was associated with mild to severe anxiety symptoms. A higher level of academic difficulty may create greater stress and anxiety for students; subsequent unsatisfactory academic results may complete the vicious cycle. Several tips were suggested by the Anxiety and Depression Association of America (ADAA) regarding test anxiety reduction.23 Female gender was not associated with mild to severe anxiety symptoms, which conflicts with previous studies.15 No clear causality was identified, but some potential aetiologies include higher stress levels from academic and interpersonal issues.

Satisfaction with friendship, sleep quality, and self-confidence were three factors inversely associated with mild to severe anxiety symptoms. Having good interpersonal relationships with peers could be a protective factor against anxiety; however, symptoms of anxiety may also hinder friendship, thus explaining the inverse relationship. Insomnia is a symptom of anxiety. According to a survey conducted by ADAA, worrying about falling asleep at night was identified as a cause for an increased level of anxiety.24 Recommendations that could help in falling asleep and achieving better sleep quality include maintaining 7 to 9 hours of uninterrupted sleep, establishing a regular bedtime routine, and avoiding electronic gadgets, coffee, and tea, as well as designing a relaxing environment for sleeping.24 Lack of self-confidence could be a cause of anxiety, but anxiety symptoms could also diminish one’s self-confidence. A bilateral association is possible.

There were several limitations in our study. Firstly, as with all cross-sectional studies, it was not possible to establish causality between the identified factors and symptoms of depression and anxiety. Hence, the identified factors are regarded as associated factors, which could be either the causes or the results of depression or anxiety. To further explore the relationships between these factors and symptoms of depression and anxiety, we reviewed the available literature to provide supporting evidence. Secondly, convenience sampling was adopted, but there could be self-selection bias when inviting students to complete the questionnaires, limited by the locations and times at which questionnaire distribution was conducted. The questionnaires were distributed in the open areas of the eight UGC-funded universities in Hong Kong. Withdrawn university students affected by depression or anxiety might not have been reached or might have refused to participate in the study; hence, the results may be underestimates. Thirdly, we could not attain our target number of 150 students from each university, except from the Hong Kong Polytechnic University. Importantly, this study used screening questionnaires to assess the depressive and anxiety symptoms experienced by the students, but no clinical diagnoses were made by health care professionals. However, with limited resources, the use of validated screening questionnaires was considered a cost-effective approach to explore the situation in general and thus was used in this study. Possible directions for further studies include the implementation of stratified random sampling, expansion of the sample size of the study, and performance of a reliability test to reduce information bias. Including detailed backgrounds of students’ disciplines and faculties, as well as ensuring a diverse population to include both local and non-local students, as well as Chinese-speaking and non-Chinese-speaking students, could allow further subgroup analysis; students from different faculties may experience depressive symptoms or anxiety symptoms at different severities due to differences in curricula. In this research, participants who were screened to have depressive and/or anxiety symptoms were not notified because of the lack of identifiable personal information. Provision of the results to participants and recommendations regarding help-seeking information are suggested for future studies. This study result may not fully reflect the severity of depressive and anxiety symptoms among students, as the study method could not reach severely depressed individuals who had socially isolated themselves.

This study showed that over 50% of university students in the eight UGC-funded universities expressed some degree of depressive symptoms (68.5%) or anxiety symptoms (54.4%). Notably, 9% of these university students exhibited moderate to severe depressive symptoms and 5.8% of the studied students showed severe anxiety symptoms. Students with regular exercise, higher self-confidence, better satisfaction with academic performance, and more optimism towards the future experienced fewer depressive symptoms. Students with better satisfaction with friendship, better sleep quality, higher self-confidence, and lower levels of academic difficulty experienced fewer anxiety symptoms.

Concept or design of study: KWC Lun, CK Chan.
Acquisition of data: All authors.
Analysis or interpretation of data: All authors.
Drafting of the article: All authors.
Critical revision for important intellectual content: KWC Lun, CK Chan.

We thank Dr LM Ho, assistant IT director of the Division of Epidemiology and Biostatistics of School of Public Health of the University of Hong Kong, for his statistical support and advice in conducting this study. We also thank the School of Public Health of the University of Hong Kong, for its support in conducting this study.

All authors have disclosed no conflicts of interest. 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.

The LKS Faculty of Medicine, University of Hong Kong reimbursed project expenses under the title of Health Research Project for HKD500.

This study was reviewed and approved by the Institutional Review Board of the University of Hong Kong/Hospital Authority Hong Kong West Cluster.

1. Hong Kong Jockey Club Centre for Suicide Research and Prevention, The University of Hong Kong. Statistics. Available from: https://csrp.hku.hk/statistics/. Accessed 18 Sep 2018.

2. Centre for Health Protection, Department of Health, Hong Kong SAR Government. Depression: beyond feeling blue. Non-Communicable Diseases Watch 2012. Available from: https://www.chp.gov.hk/files/pdf/ncd_watch_sep2012.pdf. Accessed 20 Apr 2016.

3. Centre for Health Protection, Department of Health, Hong Kong SAR Government. Population Health Survey 2003/2004. Collaborative project of Department of Health and Department of Community Medicine, University of Hong Kong, Available from: https://www.chp.gov.hk/files/pdf/report_on_population_health_survey_2003_2004_en.pdf. Accessed 20 Apr 2016.

4. Wong JG, Cheung EP, Chan KK, Ma KK, Tang SW. Web-based survey of depression, anxiety and stress in first-year tertiary education students in Hong Kong. Aust N Z J Psychiatry 2006;40:777-82. Crossref

5. Lam LC, Wong CS, Wang MJ, et al. Prevalence, psychosocial correlates and service utilization of depressive and anxiety disorders in Hong Kong: the Hong Kong Mental Morbidity Survey (HKMMS). Soc Psychiatry Psychiatr Epidemiol 2015;50:1379-88. Crossref

6. Cairns KE, Yap MB, Pilkington PD, Jorm AF. Risk and protective factors for depression that adolescents can modify: a systematic review and meta-analysis of longitudinal studies. J Affect Disord 2014;169:61-75. Crossref

7. Yu X, Tam WW, Wong PT, Lam TH, Stewart SM. The Patient Health Questionnaire-9 for measuring depressive symptoms among the general population in Hong Kong. Compr Psychiatry 2012;53:95-102. Crossref

8. Wang W, Bian Q, Zhao Y, et al. Reliability and validity of the Chinese version of the Patient Health Questionnaire (PHQ-9) in the general population. Gen Hosp Psychiatry 2014;36:539-44. Crossref

9. Martin A, Rief W, Klaiberg A, Braehler E. Validity of the Brief Patient Health Questionnaire Mood Scale (PHQ-9) in the general population. Gen Hosp Psychiatry 2006;28:71-7. Crossref

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

11. Tong X, An D, McGonigal A, Park SP, Zhou D. Validation of the Generalized Anxiety Disorder-7 (GAD-7) among Chinese people with epilepsy. Epilepsy Res 2016;120:31-6. Crossref

12. Yeh YC, Yen CF, Lai CS, Huang CH, Liu KM, Huang IT. Correlations between academic achievement and anxiety and depression in medical students experiencing integrated curriculum reform. Kaohsiung J Med Sci 2007;23:379-86. Crossref

13. O’Keeffe GS, Clarke-Pearson K, Council on Communications and Media. The impact of social media on children, adolescents, and families. Pediatrics 2011;127:800-4. Crossref

14. Hall RC, Platt DE, Hall RC. Suicide risk assessment: a review of risk factors for suicide in 100 patients who made severe suicide attempts. Evaluation of suicide risk in a time of managed care. Psychosomatics 1999;40:18-27. Crossref

15. Blanco C, Rubio J, Wall M, Wang S, Jiu CJ, Kendler KS. Risk factors for anxiety disorders: common and specific effects in a national sample. Depress Anxiety 2014;31:756-64. Crossref

16. van Ameringen M. Comorbid anxiety and depression in adults: Epidemiology, clinical manifestations, and diagnosis. 15 Jun 2017. Available from: https://www.uptodate.com/contents/comorbid-anxiety-and-depression-in-adults-epidemiology-clinical-manifestations-and-diagnosis. Accessed 24 Jul 2017.

17. Muris P. Normal and Abnormal Fear and Anxiety in Children and Adolescents. London: Elsevier; 2007.

18. Richards CS, O’Hara MW, editors. The Oxford Handbook of Depression and Comorbidity. New York: Oxford University Press; 2014.Crossref

19. Buckworth J, Dishman RK. Exercise Psychology. Champaign: Human Kinetics; 2002.

20. Faulkner GE, Taylor AH, editors. Exercise, Health and Mental Health: Emerging Relationships. New York: Routledge; 2005. Crossref

21. Wang KT. Perfectionism, depression, and self-esteem: a comparison of Asian and Caucasian Americans from a collectivistic perspective [dissertation]. Pennsylvania: The Pennsylvania State University; 2007.

22. Stirling K, Toumbourou JW, Rowland B. Community factors influencing child and adolescent depression: A systematic review and meta-analysis. Aust N Z J Psychiatry 2015;49:869-86. Crossref

23. Anxiety and Depression Association of America. Test Anxiety. Available from: https://www.adaa.org/living-with-anxiety/children/test-anxiety. Accessed 23 Mar 2017.

24. Anxiety and Depression Association of America. Stress and Anxiety Interferes with sleep. Available from: https://www.adaa.org/understanding-anxiety/related-illnesses/other-related-conditions/stress/stress-and-anxietyinterfere. Accessed 23 Mar 2017.

Intellectual disability (ID) is estimated to affect 1% to 3% of the population in Western societies.1 It is almost two-fold greater in prevalence in low-and middle-income countries, compared with high-income countries. Importantly, the General Household Survey in 2014 showed the prevalence rate of ID to be approximately 1.0% to 1.4% in Hong Kong.2

Intellectual disability is defined as ‘significant limitations both in intellectual functioning and in adaptive behaviour, as expressed in conceptual, social, and practical adaptive skills’.3 These difficulties are evident above the age of 18; ID is indicated by an intelligence quotient (IQ) of approximately two standard deviations (SD) or more below the population mean (IQ ≤70) on the IQ test.

Developmental delay (DD) describes the developmental level of a child, typically <5 years old, who is substantially below the average standard of his peers. Global DD is defined as a significant delay in two or more domains: gross motor, fine motor, language, cognitive, social, or activities of daily living. Significant delay refers to scores >2 SD below the mean on norm-referenced age-appropriate developmental tests.4

In 2015 and 2016, more than 3000 children per year were diagnosed with DD by the Child Assessment Service (CAS). A study conducted in 2003 to 2004 showed that 80% of children with significant delay and 30% of children with borderline delay were later confirmed to exhibit ID at an older age.5 Overall, 30% of children with ID had a co-morbid diagnosis of autism spectrum disorder (ASD).

The aetiology of ID is complex. While milder forms of ID are suspected to typically result from the interplay of genetic and environmental factors,6 biological causes, particularly genetic causes, are often identified in children with significant cognitive delays (IQ <50).4 Rauch et al7 studied 670 subjects, generally <6 years of age, with ID: in 39.5%, ID was related to a genetic cause; in 1.3%, it was due to an acquired or environmental cause; and in 50% to 60%, it did not exhibit a known aetiology.

Chromosomal microarray (CMA) or array comparative genomic hybridisation, is recommended by many international professional organisations as a first-tier genetic investigation for children with unexplained DD, ID, or ASD.4 8 9 Compared with conventional karyotyping, CMA is able to detect copy number variants (CNVs) with much finer resolution and is not reliant on staining and visual resolution limits. In 2010, a review of 33 published studies—involving 21 698 patients with DD, congenital anomalies, or autism—found the diagnostic yield of CMA to be 15% to 20% across all studies, compared with 3% for the standard G-banded karyotype.9 In a group of 94 patients with no symptoms other than ID, and no clear dysmorphic features, the diagnostic yield was 6.4%.8 According to the American Academy of Neurology guideline in 2011, CMA testing was abnormal in approximately 7.8% of patients with global DD or ID. The yield was higher (10.6%) in those with syndromic features.8

Children with ASD who had co-morbid ID were more likely to yield molecular diagnoses.10 Approximately 10% of patients with ASD exhibit a de-novo CNV, as detected by CMA.11 Among ASD children without syndromic features, only 6% received a molecular diagnosis.

In Hong Kong, two studies have investigated the use of CMA in patients with DD, ID, ASD, or multiple congenital anomalies (MCA). Chong et al12 found clinically significant CMA results in 20 of 105 patients (19%). Tao et al13 found a diagnostic yield of 11% for pathogenic or likely pathogenic results in 327 children, ages 1 month to >20 years. Excluding patients with MCA, the diagnostic yield of CMA for DD, ID, or ASD was approximately 4.2%.13

Chromosomal microarray has high clinical utility. Firstly, it shortens the diagnostic odyssey and may avoid unnecessary investigations, which reduces both individual and societal costs associated with testing and medical care.8 14 Secondly, it may lead to a clinically actionable recommendation. The prognostic information related to diagnosis from CMA may alert other potential co-morbid conditions that cannot be predicted on the basis of physical examination alone. In a retrospective review of 1792 patients with DD, ID, ASD, or MCA who underwent CMA testing, individuals with a positive diagnosis had a higher rate of clinical actionable recommendations than those with an uncertain result (54% vs 34%, P=0.01).15 In Hong Kong, a detection rate of 8.6% was reported for clinically actionable CNVs,13 which was comparable to the reported rates of 3.6% to 7% in Western studies.15 16 17 Thirdly, it allows estimation of recurrence risk and informed decisions regarding reproductive options for the parents’ future pregnancies.

Children’s cognitive development at ≥5 years of age is more stable if the level of ID is known. Children with DD undergo assessment at CAS to facilitate their transition into primary school. Since 2012, the Clinical Genetic Service (CGS) of the Department of Health has provided CMA testing for DD, ID, or ASD. The presence of dysmorphic features, early onset of DD, increased severity of DD, and family history are common features that prompt a genetic referral. Collaboration between CAS and CGS can potentially narrow the service gap for children with DD, ID, or ASD by enabling early access to diagnostic genomic testing, thus facilitating shorter waiting time for genetic and genomic investigation(s) and a more client-friendly turnover time for results.

The aim of this study was to investigate the genetic profile and diagnostic yield of CMA in children with moderate, severe, and profound ID. The data obtained from this study are expected to be useful in future service planning for children with DD or ID. The diagnostic yield of CMA for children with more severe forms of ID is suspected to be higher than the generally quoted figures of CMA (10%) for investigation of ID. This study targeted children with more significant ID, which is more likely to be related to an underlying genetic aetiology.

This cross-sectional territory-wide study recruited children who attended CAS for developmental assessments, before Primary 1 entry, over a 12-month period from July 2016 to June 2017. All children were at least 5 years of age. Inclusion criteria were children with moderate, severe, or profound ID, with or without ASD. According to the International Statistical Classification of Diseases and Related Health Problems, tenth revision,18 moderate ID was defined as IQ 35 to 49; severe ID was defined as IQ 20 to 34; and profound IQ was defined as IQ <20. Exclusion criteria included known causes of ID: (i) antenatal causes such as congenital brain malformation or intrauterine infections; (ii) perinatal causes such as prematurity (<34 weeks), birth asphyxia, or hypoxic ischaemic encephalopathy; (iii) postnatal causes such as intracranial bleeding, intracranial infection, or brain trauma; and (iv) other biological causes such as inborn errors of metabolism, brain tumour, neuromuscular disorders, neurodegenerative disorders, or cerebral palsy.

Unexplained ID in this study was defined as children with no identifiable causes for ID who did not meet any of the exclusion criteria. These children were non-syndromic and non-dysmorphic. In addition, they had neither MCA nor family history of ID or ASD among first- and second-degree relatives. The presence of MCA was defined as the involvement of two or more organ systems.

Children were assessed by paediatricians with the Griffiths Mental Developmental Scales,19 or by clinical psychologists with the Wechsler Preschool and Primary Scale of Intelligence–Revised.20 The Diagnostic and Statistical Manual of Mental Disorders, 4th edition,21 ASD diagnostic criteria were used for assessment of ASD.

For children with genetic diagnoses or who were known clients of CGS, their medical files were retrieved from CAS and CGS for review. Children with syndromic or dysmorphic features, MCA, or significant family history, who had not been previously referred to CGS, were referred for a formal genetic consultation before genetic or genomic investigations were recommended by a clinical geneticist.

An expedited pathway was offered for children with unexplained ID. Pre-genetic counselling was provided by a paediatrician at CAS, followed by direct blood examination for CMA and Fragile X syndrome (FGX) testing at CGS. Consultation with a geneticist was arranged if either CMA or Fragile X testing yielded abnormal outcomes. Otherwise, clients did not consult a geneticist for further counselling.

For each patient, 3 mL of blood in ethylenediaminetetraacetic acid was sent to the laboratory at CGS. All samples were tested by PerkinElmer CGX^TM v2 60K arrays designed by Agilent SurePrint technology, in accordance with the manufacturer’s instructions. The coverage of the array demonstrated an average resolution of 140 kb across the genome, and ≤40 kb in regions of clinical relevance. It evaluated >245 known genetic syndromes and >980 gene regions of functional significance in human development. Data were analysed by Genoglyphix software (Signature Genomics, Spokane [WA], United States). Genomic coordinates were based on genome assembly hg19.

Detected CNVs were systematically evaluated for clinical significance by comparison with information in the proprietary Genoglyphix Chromosome Aberration Database (Signature Genomics), internal laboratory database at CGS and the Department of Health, and public databases (Database of Genomic Variants, International Standards for Cytogenomic Arrays Consortium, and Database of Chromosomal Imbalance and Phenotype in Humans using Ensembl Resources). Categorisation of CNVs was based on available phenotypes and comparison of phenotypes with genes in the region of copy gain or loss. This was performed by searching the following databases: Online Mendelian Inheritance in Man, PubMed, RefSeq, and the University of California Santa Cruz genome browser.22 Confirmatory fluorescence in situ hybridisation (FISH), multiplex ligation-dependent probe amplification (MLPA), or conventional karyotyping was performed as indicated. Parental testing was offered to aid further interpretation and classification. Copy number variants were classified as pathogenic, likely pathogenic, uncertain clinical significance, or benign, in accordance with the 2011 American College of Medical Genetics practice guidelines.23 Only pathogenic and likely pathogenic CNVs were regarded as clinically significant.

The number of subjects to be recruited was estimated based on the average number of children with moderate, severe, and profound ID in the CAS database. In 2013 to 2015, the average number of children with moderate ID or worse was approximately 270 children per year. With the assumption that 60% of cases were unexplained,7 potential cases eligible for CMA were estimated as 160 children per year. Literature showed that the diagnostic yield of CMA was 10% in identifying abnormal cases in similar settings.13 A 95% confidence interval was desired, with a reliability (d) of 0.05, in obtaining a diagnostic yield (ˆp) of 10% in this study. The sample size needed was determined following a previously published method1:

Hence, a target sample size of 138 was needed.24

The genetic profile of children in the study was described. The diagnostic yield from CMA was calculated according to the severity of ID. The Freeman-Halton test was used to test associations between the severity of (a) ID and CMA and (b) dysmorphism and CMA findings. The null hypothesis was that there was no association between the severity of ID or dysmorphism and CMA findings. P<0.05 indicated an association between the severity of ID or dysmorphism and CMA findings. The Freeman-Halton test was conducted by using SAS/STAT 9.22.

From July 2016 to June 2017, there were a total of 339 children diagnosed with more severe forms of ID: 241 (71%) children had moderate ID, 49 (14.5%) had severe ID, and 49 (14.5%) had profound ID. Eighty-three children were excluded for the following reasons: (1) they met predefined exclusion criteria; (2) their family could not participate due to geographical reasons (eg, family lived in China); (3) a language barrier affected their understanding of study details (eg, the children or their families spoke primarily Nepalese or Sri Lankan); or (4) their parents could not be contacted for consent. A total of 31 children opted not to participate in the study. In all, 225 (66.4%) of 339 children participated in the study.

Among the 225 children, 116 (51.6%) had a co-morbid diagnosis of ASD. Male (n=151) to female (n=74) ratio was 2:1. The age ranged from 5 to 10 years old with a mean age of 6.6 years old. In all, 71.5% of children had moderate ID, 14.7% had severe ID, and 13.8% had profound ID. Two hundred twenty-one (98%) children had Chinese parents. There were two pairs of consanguineous parents: one Indian couple and one Pakistani couple.

As shown in Table 1, 68 (30.2%) children were diagnosed with a genetic condition. The percentage of a positive genetic diagnosis increased with the severity of ID. Chromosomal abnormalities comprised 76% (n=52) of the total genetic diagnoses. The most common syndromic diagnosis was Down syndrome (n=22). There were two cases of FGX. Three children had chromosome 22 microdeletion syndromes—one exhibited the more common chromosome 22q11.2 microdeletion syndrome (DiGeorge syndrome), whereas the other two exhibited chromosome 22q13.3 deletion syndrome.

Of the 225 participating children, 138 underwent CMA testing; 53 (38%) children were referred to the Clinical Genetic Service by the expedited referral pathway. Table 2 shows that 16 (11.6%) children demonstrated clinically significant CNVs that explained their ID phenotype and 10 (7.2%) had variants of uncertain significance (VUS). The diagnostic yield of CMA increased with severity of ID: it was 8.7% in moderate ID, 17.6% in severe ID, and 23.5% in profound ID (P<0.05; Table 3). Among the 16 children with clinically significant CNVs, 11 demonstrated copy number loss (deletion), four demonstrated copy number gain (duplication), and one demonstrated an unbalanced translocation between chromosome 7q and 20p (Table 4). One case of Angelman syndrome was detected by CMA and later confirmed with MLPA. One case of Cri du chat syndrome was detected by CMA and later confirmed with FISH. In total, 69% of pathogenic or likely pathogenic CNVs were de novo. Ten children (7.2%) were incidentally identified as carriers of disease: six were alpha thalassemia heterozygous carriers, one was a heterozygous carrier of Joubert syndrome type 4, one was a heterozygous carrier of autosomal recessive disease Joubert syndrome and nephronophthisis, one was a heterozygous carrier of autosomal recessive deafness affecting the OTOA gene, and one was a carrier of Klinefelter syndrome.

The overall diagnostic yield of CMA among children with ID (11.6%) was consistent with studies performed in other regions of the world. The diagnostic yield of CMA increased with severity of ID and was much higher in children with dysmorphism (45.8% vs 4.4%, P<0.05).

Variants of uncertain significance are not uncommon. In all, 7.2% of children in this cohort had VUS (Table 5). Regular follow-up and reassessment by a clinical geneticist is necessary for these children, because VUS may eventually be re-classified as pathogenic or benign as clinical and genomic data accumulate in the literature.

The paradigm shift in the medical genetic and genomic field from the phenotype-first approach to the genotype-first approach is revolutionary. Traditionally, a phenotype-first approach was used to guide the investigation of possible genetic diagnoses, eg, karyotyping for Down syndrome, or specific assays, such as FISH, for DiGeorge syndrome. In the past decade, CMA has allowed more comprehensive unbiased discovery of microdeletion and microduplication syndromes throughout the human genome. Since the 1980s, it has been well-known that certain chromosomal microdeletion and microduplication syndromes are associated with recognisable forms of ID and DD. Classical examples include 15q11-q13 deletion, associated with Prader-Willi and Angelman syndromes, and 22q11.2 deletion, associated with DiGeorge syndrome (velocardiofacial syndrome). Thus far, approximately 50 to 60 recurrent microdeletion or duplication syndromes have been identified in children with DD or ID.

Although CMA is robust, it cannot replace a formal genetic consultation for children with clinically suspected genetic conditions. As an example, in Prader-Willi syndrome, 70% to 75% of cases can be detected by CMA, as they are due to a paternal 15q microdeletion subtype; 20% to 25% of cases require a more specific methodology for genetic confirmation. Therefore, clinical correlation and expert assessment remain necessary.

Males are more susceptible to ID than females; more than 100 X-linked genes are associated with ID.25 X-linked ID constitutes 5% to 10% of ID in males. One of the best-known causative genes for ID is FMR1; mutations of FMR1 result in FGX. The estimated incidence of FGX is approximately 1 in 4000 males and 1 in 5000 to 1 in 8000 females (approximately 0.5% of cases of ID) in Western countries. Peprah26 reported that the incidence of FGX in countries/regions with significant Asian populations, such as Canada, Estonia, Japan, and Taiwan, was significantly lower than in Western countries. In a study of 553 male children between the ages of 6 months and 18 years, Chen et al27 estimated the prevalence of FGX in mainland China to be approximately 0.93% among children with moderate to severe ID. Among the 225 children in our cohort, only two were diagnosed with FGX. Both exhibited ASD and severe ID. The typical physical characteristics of FGX, such as narrow face, protruding ears, and macro-orchidism, are often less obvious in early childhood; notably, they may become more prominent as the child approaches adolescence. This lack of early physical characteristics increases the diagnostic challenge for clinicians. Fragile X syndrome testing, regarded as first-tier genetic testing for DD and ASD in many international guidelines, has been a standard genetic investigation for ID or ASD in Hong Kong for many years.

Incomplete penetrance of a genomic condition within the same family is not uncommon. Notably, there were two such cases of 16p13.11 microduplication syndrome in this cohort. Patient 12 (Table 4) exhibited subtle dysmorphism comprising downslanting palpebral fissures, prominent ears, and mild right ptosis. Left undescended testes and umbilical hernia were operated in infancy. He exhibited global DD and was later diagnosed with moderate ID. His mother and two sisters had an identical chromosomal defect, but exhibited normal intelligence. Patient 13 demonstrated a more severe phenotype with hirsutism, bushy eyebrows, frontal bossing, hearing loss, visual problems, and profound ID. His mother and elder brother, both carrying the microduplication, exhibited normal intelligence. There likely exist unknown environmental or genetic modifiers to modulate susceptibility to ID caused by this microduplication. Thus, relying on family history to determine whether ID is hereditary can be misleading.

An important aspect with respect to obtaining a genetic diagnosis is patient prognosis. The 16p13.11 duplication syndrome is associated with an aortic root defect. In this study, the two affected children and their affected family members were referred for monitoring by echocardiogram. Similarly, Patient 2 exhibited 3q13.31 microdeletion syndrome, which is associated with diabetes mellitus and deafness; this patient was referred for audiological assessment and counselled on lifestyle management to minimise the risk of diabetes. In this study, three of 16 CMA-positive cases (18.8%) were clinically actionable.

Pre-test genetic counselling is as important as post-test counselling. Coincidental findings of genetic changes that either predict adult-onset conditions or reveal carrier status for recessive or X-linked conditions are common. In the present cohort, 10 children were identified as carriers of genetic conditions, including one child diagnosed with Klinefelter syndrome. He presented with moderate ID and ASD. Chromosomal microarray identified a copy number gain of the entire X chromosome. Klinefelter syndrome can be associated with learning disabilities, as well as delayed speech and language development. While a small, but significant, downward shift in mean overall IQ has been reported, general cognitive abilities of patients with Klinefelter syndrome are not typically in the ID range.28 An extra X chromosome may have contributed partially, but could not entirely explain the severity of ID. The major implications are that individuals with Klinefelter syndrome have a higher risk of endocrine dysfunction, fertility problems, male breast cancer, and autoimmune disease.

This study provided important information with respect to service planning for children with ID in Hong Kong. It allowed testing of an expedited referral mechanism between CAS and CGS, in which cases with unexplained ID benefitted through a significant reduction of waiting time for both pre-testing genetic counselling and investigation turnover time. This study included 38% of the 138 children who were not referred to CGS. The ideal future approach may be to extend the expedited mechanism for children with early-onset significant DD. It can avoid unnecessary investigations, thus lowering stress for both child and parent; importantly, it may reduce societal costs.

There were several limitations in this study. Firstly, a complete genetic profile of ID was not generated, as this cohort excluded mild ID. Secondly, clients from minority cultural groups in Hong Kong were underrepresented, because the language barrier affected recruitment. More effort must be expended to ensure equal opportunities for children from diverse cultural backgrounds. Thirdly, the duration of the study was of insufficient length for commentary on trends regarding the genetic profile of ID in Hong Kong.

The overall diagnostic yield (11.6%) of CMA is compatible with other international cohorts. Chromosomal microarray yield increases with the severity of ID. These data further support the use of CMA as a first-tier investigation for children with significant unexplained ID in Hong Kong.

All authors have made substantial contributions to the concept or design of the study, acquisition of data, analysis or interpretation of data, drafting of the article, and critical revision for important intellectual content.

The authors would like to thank Mr Morris Wu of Child Assessment Service, Department of Health for his advice on the statistical analysis of the data.

The authors have no conflicts of interest to disclose. 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.

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

Approval was obtained from the Ethics Committee of the Department of Health, Hong Kong Special Administrative Region. Informed consent was obtained from parents or legal guardians. Parents and legal guardians were counselled about the indication for CMA, benefits and limitations of test, methodology, reporting time, and possible outcomes upon recruitment.

1. Leonard H, Wen X. The epidemiology of mental retardation: challenges and opportunities in the new millennium. Ment Retard Dev Disabil Res Rev 2002;8:117-34. Crossref

2. Social Data Collected via the General Household Survey. Special Topics Report No 62. Persons with Disabilities and Chronic Diseases. Hong Kong: Census and Statistics Department, Hong Kong SAR Government; 2014.

3. American Association on Intellectual and Developmental Disabilities. Intellectual Disability: Definition, Classification, and Systems of Supports. 11th ed. Washington, DC: American Association on Intellectual and Developmental Disabilities;2010: xvi, 259.

4. Moeschler JB, Shevell M, Committee on Genetics. Comprehensive evaluation of the child with intellectual disability or global developmental delays. Pediatrics 2014;134:e903-18. Crossref

5. Tang KM, Chen TY, Lau VW, Wu MM. Clinical profile of young children with mental retardation and developmental delay in Hong Kong. Hong Kong Med J 2008;14:97-102.

6. Willemsen MH, Kleefstra T. Making headway with genetic diagnostics of intellectual disabilities. Clin Genet 2014;85:101-10. Crossref

7. Rauch A, Hoyer J, Guth S, et al. Diagnostic yield of various genetic approaches in patients with unexplained developmental delay or mental retardation. Am J Med Genet A 2006;140:2063-74. Crossref

8. Michelson DJ, Shevell MI, Sherr EH, Moeschler JB, Gropman AL, Ashwal S. Evidence report: Genetic and metabolic testing on children with global developmental delay: report of the Quality Standards Subcommittee of the American Academy of Neurology and the Practice Committee of the Child Neurology Society. Neurology 2011;77:1629-35. Crossref

9. Miller DT, Adam MP, Aradhya S, et al. Consensus statement: chromosomal microarray is a first-tier clinical diagnostic test for individuals with developmental disabilities or congenital anomalies. Am J Hum Genet 2010;86:749-64. Crossref

10. Tammimies K, Marshall CR, Walker S, et al. Molecular diagnostic yield of chromosomal microarray analysis and whole-exome sequencing in children with autism spectrum disorder. JAMA 2015;314:895-903. Crossref

11. Sebat J, Lakshmi B, Malhotra D, et al. Strong association of de novo copy number mutations with autism. Science 2007;316:445-9. Crossref

12. Chong WW, Lo IF, Lam ST, et al. Performance of chromosomal microarray for patients with intellectual disabilities/developmental delay, autism, and multiple congenital anomalies in a Chinese cohort. Mol Cytogenet 2014;7:34. Crossref

13. Tao VQ, Chan KY, Chu YW, et al. The clinical impact of chromosomal microarray on paediatric care in Hong Kong. PLoS One 2014;9:e109629. Crossref

14. Tirosh E, Jaffe M. Global developmental delay and mental retardation—a pediatric perspective. Dev Disabil Res Rev 2011;17:85-92. Crossref

15. Coulter ME, Miller DT, Harris DJ, et al. Chromosomal microarray testing influences medical management. Genet Med 2011;13:770-6. Crossref

16. Ellison JW, Ravnan JB, Rosenfeld JA, et al. Clinical utility of chromosomal microarray analysis. Pediatrics 2012;130:e1085-95. Crossref

17. Riggs ER, Wain KE, Riethmaier D, et al. Chromosomal microarray impacts clinical management. Clin Genet 2014;85:147-53. Crossref

18. World Health Organization. The ICD-10 classification of mental and behavioural disorders: diagnostic criteria for research: Geneva: World Health Organization; 1993.

19. Luiz D, Barnard A, Knosen N, Kotras N, Faragher B, Burns LE. Griffiths Mental Development Scales–Extended Revised. Two to Eight Years. Administration Manual. Oxford, UK: Hogrefe; 2006.

20. Wechsler D. Wechsler Preschool and Primary Scale of Intelligence–Revised. WPPSI-R: Psychological Corporation; 1989.

21. Diagnostic and Statistical Manual of Mental Disorders: DSM-IV. Washington, DC: American Psychiatric Association;1994: 535.

22. Kan AS, Lau ET, Tang WF, et al. Whole-genome array CGH evaluation for replacing prenatal karyotyping in Hong Kong. PLoS One 2014;9:e87988. Crossref

23. Kearney HM, Thorland EC, Brown KK, et al. American College of Medical Genetics standards and guidelines for interpretation and reporting of postnatal constitutional copy number variants. Genet Med 2011;13:680-5. Crossref

24. Boada R, Janusz J, Hutaff-Lee C, Tartaglia N. The cognitive phenotype in Klinefelter syndrome: a review of the literature including genetic and hormonal factors. Dev Disabil Res Rev 2009;15:284-94. Crossref

25. Lubs HA, Stevenson RE, Schwartz CE. Fragile X and Xlinked intellectual disability: four decades of discovery. Am J Hum Genet 2012;90:579-90. Crossref

26. Peprah E. Fragile X syndrome: the FMR1 CGG repeat distribution among world populations. Ann Hum Genet 2012;76:178-91. Crossref

27. Chen X, Wang J, Xie H, et al. Fragile X syndrome screening in Chinese children with unknown intellectual developmental disorder. BMC Pediatr 2015;15:77. Crossref

28. Daniel WW. Biostatistics: A Foundation for Analysis in the Health Sciences. 9th ed. Hoboken, NJ: J Wiley & Sons; 2009.

Telepsychiatry is the practice of delivering mental health consultations at a distance. New developments in information and communication technologies have allowed telepsychiatry to become a viable method of providing services to patients in rural or remote locations with limited access to medical services.1 2 3 Telepsychiatry has been used in prison settings for more than 20 years.4 A demonstration of telepsychiatry in prison in the US in 1996 concluded that this practice was cost-effective.5 Prison administrators even claimed that there were fewer assault incidents after its use.5 In Hong Kong, the use of telepsychiatry can be dated back to 1998.6

Currently, persons in custody (PICs) in Hong Kong receive primary medical care within the institutions of the Correctional Services Department (CSD). However, for specialist psychiatric service for their mental problems, PICs must attend psychiatric specialist out-patient clinics (SOPCs) of the Hospital Authority. In addition, most psychiatric drugs are available only in SOPCs.

For security reasons, PICs must be escorted by at least two CSD staff and be handcuffed on every occasion they need to attend follow-up at a SOPC. Such an exposing arrangement inevitably causes much embarrassment and stigmatisation for PICs. There is also a potential risk to the public in the event of abscondence from custody. The PIC may also experience travel sickness during the journey between the correctional facilities and the SOPC; most correctional facilities in Hong Kong are situated in relatively remote areas, so the journey times can be long. Furthermore, other patients in the SOPC may feel uncomfortable witnessing a PIC being handcuffed. Some SOPCs manage this problem by placing the PIC in a special corner or room, depending on availability.

Face-to-face consultation is the gold standard for medical practice. However, telepsychiatry is suitable for PICs and might confer additional benefits for this group of patients. Direct physical examinations are typically unnecessary for stable psychiatric patients during follow-up. Additionally, the nurse at the CSD site can help measure vital signs such as blood pressure, pulse, and temperature. Therefore, offering PICs psychiatric teleconsultations cannot only maintain their usual psychiatric care but also reduce embarrassment, stigmatisation, and the risk of abscondence. Furthermore, it can also reduce the need for special arrangements in SOPCs.

To the best of our knowledge, there have been no previous studies exploring the effect of psychiatric teleconsultations for Chinese psychiatric out-patients under the legal custody of the CSD in Hong Kong. The main aim of the present study was to explore the use of psychiatric teleconsultations for stable Chinese psychiatric out-patients under the legal custody of the CSD in Hong Kong. The desired outcome was to maintain the clinical interests of PICs and to provide them with appropriate psychiatric services using telecommunications in a safe, humane, and cost-effective manner. This was an interventional pilot study evaluating the effect of psychiatric teleconsultations on the general health of an intervention group of clinically stable Chinese male psychiatric out-patients who were under the custody of the CSD as compared with a matched control group of Chinese male psychiatric out-patients under the usual type of care (ie, face-to-face consultation with a psychiatrist at a SOPC). In addition, the satisfaction of patients towards psychiatric teleconsultations was assessed.

The null hypotheses of this study were as follows:
1. After the consultation, the psychological health of the intervention group is worse than that of the control group;
2. The effect of psychiatric teleconsultations is unsustainable;
3. Patients in the intervention group are unsatisfied with the psychiatric teleconsultations;
4. Adverse events occur during psychiatric teleconsultations.

This was an interventional pilot study conducted by the Hospital Authority in collaboration with the Hong Kong CSD.

The study period was from June 2014 to May 2016. Participants were aged 21 to 64 years. The intervention group comprised Chinese patients in custody attending follow-up at the SOPC of Castle Peak Hospital (CPH), Hong Kong. The control group included Chinese patients in custody attending follow-up at other SOPCs in Hong Kong. In this study, only male PICs were included because of logistic and feasibility reasons. Exclusion criteria applied when selecting intervention and control participants for this study included: (a) patients with mental instability or with prominent and recent change/deterioration in mental condition, such as those who were suicidal or homicidal, or who had delirium or acute psychosis; (b) patients who required regular blood tests, such as those taking clozapine; (c) patients requiring other tests/investigations only available in SOPC or Hospital Authority hospitals; (d) patients requiring drug administration in SOPC, such as depot antipsychotics; (e) patients attending SOPC for the first time; or (f) patients with visual or auditory deficits that might impair the ability to interact via video-conferencing. Eligible patients meeting the inclusion criteria were invited to participate in the study. Written informed consent was obtained from the intervention and control participants.

A sample size of at least 80 participants for the intervention group with an intervention-to-control ratio of approximately 1:3 was adopted in this study. This was an affordable and representative sample size with reference to the number of stable psychiatric out-patients in custody attending follow-up appointment at the SOPC of CPH during the study period.

Socio-demographic and clinical data including age, education level, and principal psychiatric diagnosis according to the 10th revision of the International Classification of Diseases were obtained.7 The intervention and control participants were requested to complete the Chinese version of the 12-item General Health Questionnaire (C-GHQ-12).8 The GHQ is a self-administered test used for evaluating the psychological components of ill health and is helpful in screening for general emotional distress.9 10 The GHQ possesses adequate content validity and construct validity, and good internal consistency has been demonstrated with Cronbach’s alphas ranging from 0.82 to 0.93.9 10 The Chinese and English versions of the GHQ have been adopted for Chinese and non-Chinese subjects, respectively.8 9 10 The C-GHQ-12 consists of 12 items, with each item assessing the severity of a mental problem using a 4-point Likert scale.8 The six positive items were rated from 1 (more than usual) to 4 (much less than usual) and the six negative items from 1 (never) to 4 (much more than usual); thus, a higher score indicates a more severe mental health condition. In this study, the pre-consultation and post-consultation C-GHQ-12 scores for each patient were obtained. The difference between the two scores (ie, the pre-post difference) was used as a proxy measurement of the quality of consultation.

The intervention participants were also requested to complete a questionnaire in Chinese designed to measure the patient satisfaction regarding the psychiatric teleconsultation. The questionnaire consisted of nine statements/questions rated according to a 5-point Likert scale, from 1 (strongly agree) to 5 (strongly disagree) or from 1 (very satisfied) to 5 (very dissatisfied). The questionnaire was designed by the authors as there were no available validated Chinese questionnaires suitable for assessing patient satisfaction of telepsychiatry at the time of the study (English translations of the statements/questions in the questionnaire are listed in Table 1).

The intervention participants were transferred from various CSD institutions to the Lai Chi Kok Reception Centre, Hong Kong, for the psychiatric teleconsultation. On the scheduled day of consultation, the CSD staff brought two portable video-conferencing devices to CPH. Registration was performed only after the device had been checked as functional. All persons in the consultation rooms at the CSD site and at the CPH site were identified to each other prior to the consultation session. Consultation rooms provided at both sites were appropriately set up with particular attention to audio and visual privacy, lighting, backdrop, and gaze angle. A qualified CSD nurse was present in the consultation room at the CSD site together with the patient. There was also a CSD medical doctor at the reception centre during the psychiatric teleconsultation, in case emergency medical treatment was needed. After the consultation, the CSD staff collected the medicine for the patient according to the usual procedures. For the intervention participants, the maximum number of consecutive psychiatric teleconsultations was set as four, after which a face-to-face follow-up consultation must follow. The control participants attended only face-to-face consultations at other SOPCs. Both groups of participants filled in the C-GHQ-12 within 7 days before the consultation and again within 7 days after the consultation. In addition, the intervention participants filled in the satisfaction survey questionnaire after the psychiatric teleconsultation. Any major adverse events, such as medical or psychiatric emergencies, were recorded.

Descriptive statistics were used to analyse the baseline profile of the participants’ socio-demographic and clinical characteristics as well as pre- and post-consultation C-GHQ-12 scores and satisfaction survey questionnaire responses. Chi squared test and two-samples t test were performed to assess if there were differences in the baseline characteristics between the intervention and control participants. A non-parametric Mann-Whitney U test was performed to test if there were differences in the pre-post difference in C-GHQ-12 score between intervention and control participants attending their first consultation. Spearman’s correlation was used to compute the correlation between the pre-post difference in C-GHQ-12 score of the first and second teleconsultations among the intervention participants. All statistical analyses were conducted using SPSS for Windows, version 12.0 (SPSS Inc, Chicago [IL], US), with P<0.05 considered as statistically significant in this study.

During the study period, there were 377 PIC scheduled attendances at CPH. Of these, 221 PIC scheduled attendances were suitable for psychiatric teleconsultation; however, for 49 of the suitable PIC scheduled attendances, the PICs refused to give consent for this study. Finally, 172 PIC psychiatric teleconsultation attendances were included. Each participant could have more than one psychiatric teleconsultation attendance during the study period. Therefore, 86 participants aged 21 to 64 years who were stable Chinese male psychiatric out-patients and who fulfilled the inclusion and exclusion criteria for psychiatric teleconsultations in CPH were included. For the control group, 249 male patients within the same age range (21-64 years) were recruited.

Table 2 compares patient characteristics between the intervention and control groups. There were no significant differences in the age and education profile between the two groups. The mean age of both groups was approximately 40 years. Approximately three quarters of the participants in each group had attained education at the secondary level or above. There was a significant difference in the principal psychiatric diagnosis (P=0.029). Slightly over 50% of each group were diagnosed as substance abuse. A larger proportion of intervention participants had schizophrenia (28%) than did the control participants (16%). There were no significant differences in the mean pre-consultation C-GHQ-12 score between the two groups.

The mean (standard deviation) on-air time duration of the psychiatric teleconsultations was 6.33 (3.58) minutes. There were no significant adverse events associated with teleconsultations reported during the study. The pre-post difference in C-GHQ-12 score of the intervention participants was significantly higher than that of the control participants (P=0.023; Table 2). Furthermore, among 29 intervention participants who had at least two teleconsultations, the association between pre-post difference in C-GHQ-12 score of the first and second teleconsultations was moderately strong (r=0.309, P=0.103) but did not reach the level of significance set for this study. The possible scores on the satisfaction survey questionnaire ranged from 9 (the most satisfied) to 45 (the least satisfied). The mean (standard deviation) satisfaction score of the intervention group was 16.48 (4.35). No major adverse events were reported throughout the study.

Telepsychiatry is not a new development in Hong Kong. Since 2001, the use of telepsychiatry has been shown to increase access to care.11 Studies have shown that telepsychiatry is an effective means to provide psychogeriatric services to residents of care homes,11 and cognitive intervention for community-dwelling elderly patients with memory problems.12 This was the first intervention pilot study in Hong Kong exploring the effect of psychiatric teleconsultation for Chinese psychiatric out-patients under the legal custody of the CSD. Our study compared the effectiveness of psychiatric teleconsultations with that of face-to-face consultations among PIC receiving out-patient psychiatric treatment. The two groups of stable Chinese male out-patient participants had the same baseline characteristics in age, education level, and pre-consultation C-GHQ-12 score. The results showed that the standard of care of teleconsultations was comparable to that of face-to-face consultations. The pre-post difference in C-GHQ-12 score for teleconsultations had a marginally larger positive increase than did face-to-face consultations. The intervention participants also showed high satisfaction with the psychiatric teleconsultation service, with a mean satisfaction score above the 80th percentile. This results is similar to that of a previous study in the US that compared the effectiveness of telepsychiatry and in-person psychiatry sessions among 71 parolees over a 6-month follow-up period, revealing high satisfaction with telepsychiatry treatment.13 In the present study, the pre-post difference between C-GHQ-12 score of the first and second psychiatric teleconsultations showed a moderate positive relationship, suggesting a consistent and sustainable clinical effect of telepsychiatry between sessions. Importantly, there were no reports of significant adverse events. Therefore, telepsychiatry can be considered sustainable and safe for Chinese PIC in Hong Kong.

Studies on the effects of telepsychiatry for incarcerated populations are relatively scarce; however, the results of our study are also consistent with studies of telepsychiatry in populations that were not involved with the correctional system. In a randomised controlled study in Canada, 495 patients were assigned at random to be examined face-to-face or by telepsychiatry.14 Psychiatric consultations and follow-ups delivered by telepsychiatry produced clinical outcomes equivalent to those achieved when the service was provided face-to-face.14 This result suggests that psychiatric consultation and short-term follow-up can be as effective when delivered by telepsychiatry as when provided face-to-face. Another study evaluating the effectiveness of telepsychiatry in relation to cognitive changes in patients with dementia revealed that changes in the Mini-Mental State Examination score were not significantly different between patients receiving teleconsultations and those receiving clinic-based face-to-face consultations.15 This finding suggest that telepsychiatry may be a useful alternative to face-to-face clinical visits for treatment of a wide range of patient groups, including patients with dementia. Research has shown that there is an association between dementia and criminal offences and that the use of telepsychiatry might be extended to PIC with dementia in Hong Kong.16

Our study has several limitations. First, the sample size of the intervention group was small. Most enrolled patients served short sentences but had long follow-up intervals, because we recruited stable patients. Although the study duration was 2 years, only 29 out of 86 intervention participants had at least two psychiatric teleconsultations for comparison within the intervention group. Second, recruitment of potential telepsychiatry participants was limited to stable Chinese male psychiatric out-patients. Therefore, the sample may be affected by self-selection bias, because PICs volunteered to participate in telepsychiatry. This restricted sample also limits the generalisability of the results. Third, the mean time between the consultation and completion of the C-GHQ-12 was not recorded for intervention or control groups. Differences in this time interval may affect the pre-post difference in C-GHQ-12 score for the groups. We also did not collect data on the follow-up interval between the first and second psychiatric teleconsultations for the intervention group. This follow-up interval may affect the C-GHQ-12 score and the satisfaction on psychiatric teleconsultation. Fourth, the mean on-air time duration of psychiatric teleconsultations was 6.33 minutes; however, we did not measure the duration of face-to-face consultations for comparison between the two groups. The duration of consultation may have an effect on the outcome scores and patient satisfaction. Last, the study lacked robust clinical outcome measures and the satisfactory questionnaire adopted in this study had not been validated. Despite the limitations of this study, the results suggest that psychiatric teleconsultations are a sustainable and safe alternative to face-to-face consultations for stable Chinese male psychiatric out-patients in custody. The use of psychiatric teleconsultations has potential in other populations of PICs, such as female PICs or elderly PICs, but further research is required to investigate psychiatric teleconsultations for these populations.

Further research is required to examine the full potential of telepsychiatry among PICs in Hong Kong. In future studies, female patients should be recruited, to assess any sex-based differences. In addition, the scale of future studies using telepsychiatry can be increased by setting up more stations at CSD institutions and at other SOPCs in the Hospital Authority. Clinical outcomes such as symptom severity and psychological functioning of the patients could be assessed. Given the increasing number of older PICs in Hong Kong, recruitment of older patients could be considered for further study. It would be worthwhile to perform a future study with a larger sample size and with participants receiving a greater number of psychiatric teleconsultations, in order to further support the sustainability of psychiatric teleconsultations.

A cost analysis for psychiatric teleconsultation was beyond the scope of the present study. However, a systematic review of 137 telemedicine services in hospital facilities revealed that one of the key reasons for introducing telemedicine was cost reduction.17 Similar cost-saving conclusions have been reported in two studies in Hong Kong related to dementia and community geriatric services.11 12 In future studies, cost analysis of psychiatric teleconsultation including direct, indirect, and hidden costs could be calculated for further exploring the effectiveness of psychiatric teleconsultation.

Telepsychiatry appears to be an acceptable approach for providing out-patient psychiatric care for stable Chinese male PICs in Hong Kong. Telepsychiatry can be considered a sustainable and safe alternative to face-to-face consultations, with a comparable standard of care. Moreover, the intrinsic problems of embarrassment and stigmatisation caused to PICs, the risk that PICs might abscond, and the safety of the general public are all addressed by this promising alternative mode of psychiatric care for stable Chinese male PICs. Telepsychiatry is likely to show similar benefits for Chinese female PICs and PICs in other age-groups, such as older adults, but further research is required to confirm this.

All authors have made substantial contributions to the concept and design, acquisition of data, analysis and interpretation of data, drafting and critical revision for important intellectual content of the article.

We would like to acknowledge Dr CK Tung, Dr CF Tsui, Mr KW Chung, Mr Kenny Wong, Dr NM Kwong, and all the mental health professionals and CSD staff who have assisted in the design and implementation of this study. We would also like to acknowledge the study participants who had kindly participated in the present study.

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

All authors have disclosed no conflicts of interest. 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. The research was presented in part in the Hospital Authority Convention 2017, 16 May 2017, Hong Kong.

Approval for conducting the study was granted by the Research and Ethics Committee of the New Territories West Cluster of the Hospital Authority and the Research and Ethics Committee of the Correctional Services Department. The principles outlined in the Declaration of Helsinki were followed in the conduction of this study.

1. Gray LC, Edirippulige S, Smith AC, et al. Telehealth for nursing homes: the utilization of specialist services for residential care. J Telemed Telecare 2012;18:142-6. Crossref

2. Grady B, Myers KM, Nelson EL, et al. Evidence-based practice for telemental health. Telemed e-Health 2011;17:131-48. Crossref

3. WHO Global Observatory for eHealth. 2010. Telemedicine: opportunities and developments in Member States: report on the second global survey on eHealth. Geneva: World Health Organization. Available from: http://www.who.int/iris/handle/10665/44497. Accessed 1 Jan 2018.

4. McLaren P. Telemedicine and telecare: what can it offer mental health services? Adv Psychiatr Treat 2003;9:54-61. Crossref

5. Telemedicine can Reduce Correctional Health Care Costs: An Evaluation of a Prison Telemedicine Network. US Department of Justice; 1999.

6. Hjelm NM. Telemedicine: academic and professional aspects. Hong Kong Med J 1998;4:289-92.

7. International Statistical Classification of Diseases and Related Health Problems, 10th revision. Geneva, World Health Organization; 1992.

8. Chan DW. The Chinese version of the General Health Questionnaire: does language make a difference? Psychol Med 1985;15:147-55. Crossref

9. Goldberg DP. The Detection of Psychiatric Illness by Questionnaire. London: Oxford University Press; 1972.

10. Goldberg D, Williams P. A User’s Guide to the General Health Questionnaire. London: NFER; 1988.

11. Tang WK, Chiu H, Woo J, Hjelm M, Hui E. Telepsychiatry in psychogeriatric service: a pilot study. Int J Geriatr Psychiatry 2001;16:88-93. Crossref

12. Poon P, Hui E, Dai D, Kwok T, Woo J. Cognitive intervention for community-dwelling older persons with memory problems: telemedicine versus face-to-face treatment. Int J Geriatr Psychiatry 2005;20:285-6. Crossref

13. Farabee D, Calhoun S, Veliz R. An experimental comparison of telepsychiatry and conventional psychiatry for parolees. Psychiatr Serv 2016;67:562-5. Crossref

14. O’Reilly R, Bishop J, Maddox K, Hutchinson L, Fisman M, Takhar J. Is telepsychiatry equivalent to face-to-face psychiatry? Results from a randomized controlled equivalence trial. Psychiatr Serv 2007;58:836-43. Crossref

15. Kim H, Jhoo JH, Jang JW. The effect of telemedicine on cognitive decline in patients with dementia. J Telemed Telecare 2017;23:149-54. Crossref

16. Liljegren M, Naasan G, Temlett J, et al. Criminal behavior in frontotemporal dementia and Alzheimer disease. JAMA Neurol 2015;72:295-300. Crossref

17. AlDossary S, Martin-Khan MG, Bradford NK, Smith AC. A systematic review of the methodologies used to evaluate telemedicine service initiatives in hospital facilities. Int J Med Inform 2017;97:171-94. Crossref

Over the past 30 years, the proportion of the Hong Kong population aged ≥65 years doubled from 6.6% in 1981 to 13.3% in 2011.1 The proportion of those aged ≥65 years among patients with tuberculosis (TB) tripled from 13%2 to 39%3 in the same period. Although the annual notification rate decreased from 149.1 to 65 per 100 000 population and the TB mortality rate decreased from 9.4 to 2.6 per 100 000 population from 1981 to 2011, the proportion of those aged ≥65 years increased from 53% to 82% among TB deaths.2 3 In older adults, TB is associated with other co-morbidities, hospitalisation, and delays in presentation and commencement of treatment.4 Missed opportunities for intervention might contribute to the higher mortality rates in older adults, and might also increase the risk of TB transmission. The present longitudinal study was conducted to assess the effects of age on the mortality rates of patients with TB and to elucidate the factors associated with missed TB diagnosis.

All consecutive cases of TB notified to the Department of Health in 2010 were retrospectively collected from the statutory TB notification registry. Hong Kong identity card numbers (or passport numbers for non-residents) were retrieved from the notification registry, together with date of notification, source of notification, and demographic and clinical information. Further clinical information and outcome data at 1 year after notification/initiation of treatment were retrieved from the TB programme record forms.3 These forms are filed by the TB and Chest Service for patients managed under its chest clinics and for patients managed by other health care providers. Treatment outcome was classified according to the World Health Organization (WHO) recommendations.5 Using the identity card number/passport number as the unique identifier, the 2010 TB cohort data were cross-matched with the statutory death registry from 1 January 2009 till 31 December 2012 for vital status, date and cause(s) of death. All cases with a date of TB notification after the date of death were recorded. A mortality survey was conducted on these recorded cases by retrieving relevant clinical information from records in public clinics and hospitals.

The demographics, co-morbidities, treatment outcomes, and mortality pattern of the cohort were analysed. Published data on patients of all ages with TB6 and on elderly patients with TB7 notified in 1996 were used for comparison. A date of TB notification after the date of death was considered as a surrogate marker of delayed diagnosis. Categorical variables were analysed by Pearson χ² test or Fisher’s exact test as appropriate; continuous variables were analysed by Mann-Whitney U test. Binary regression modelling was used to calculate the adjusted odds ratios (aORs) for risk factors for delayed diagnosis of TB after death using a backward conditional approach, with probability to remove being 0.10 and to retain being 0.05. A two-tailed P<0.05 was considered statistically significant. Statistical analyses were performed using SPSS for Windows, version 16.0 (SPSS Inc, Chicago [IL], US).

After exclusion of 336 cases subsequently denotified because of alternative diagnoses, a total of 5092 patients with TB were included in the 2010 TB cohort, at a notification rate of 72.5/100 000 person-years. Table 1 summarises their demographic data, clinical characteristics, and 1-year outcomes. Comparison with published data on the 1996 TB cohort6 7 was restricted to patients managed under the TB and Chest Service, for which the proportion of patients with TB aged ≥60 years increased from 34.5% in 1996 to 42.9% in 2010. There were more co-morbidities such as diabetes mellitus (16.0% vs 9.6%), lung cancer (2.1% vs 1.1%), and other cancers (5.0% vs 0.6%) in the 2010 TB cohort than in the 1996 TB cohort (χ² test, P<0.001). In 2010, the proportion of patients who died before completion of TB treatment was smaller for those managed under the TB and Chest Service (7.4%) than for the overall cohort (16.2%). However, the proportion of patients managed under the TB and Chest Service who died before completion of TB treatment nearly doubled between 1996 (3.9%) and 2010 (7.4%).

Among 5092 TB notifications, 1061 (20.9%) deaths occurred within 2 years of notification. Of the 1061 deaths, 211 (4.1%) occurred before the TB notification (ie, TB was notified after death; median delay in notification [interval between death and TB notification] 45 days, interquartile range 30-65 days). Of the deaths after notification, 683 (13.4%) died in the first year and 167 (3.3%) died in the second year. The reported causes of death were related to TB in only 191 (18.0%) of all deaths; 30 (14.2%) before TB notification, 158 (23.1%) in the first year, and three (1.8%) in the second year after notification.

Among the 211 deaths before TB notification, only 30 (14.2%) had TB as the main cause of death. There were 54 cases of ‘pneumonia unspecified’ and three cases of ‘sepsis unspecified’ reported as main cause of death. Of these cases, only five with potential causative organisms, such as Pseudomonas, Acinetobacter, or Escherichia coli, were identified. However, in the sputum that had been collected before death in these patients, Mycobacterium tuberculosis was subsequently isolated after prolonged culture, indicating that TB was the likely main cause of death in the remaining 52 deaths initially reported as ‘sepsis or pneumonia unspecified’. Including these revised results increases the 2010 TB-related mortality from 191 to 243, ie, an increase of 27% from the officially reported mortality figures of 2.6 to 3.4 per 100 000 person-years.8 The corresponding proportion of TB-related mortality increases to 38.8% (82/211) in cases with TB notified after death compared with 23.3% (158/683) who died in the first year after notification and 1.7% (3/167) who died in the second year. Therefore, a substantial proportion (15.5%) of TB-related deaths could potentially have been prevented by early diagnosis and treatment.

For the 211 deaths before TB notification, 25 cases of TB were notified from the public mortuary. Of the remaining 186 cases of TB that were notified from hospital, 173 hospital records were collected; 13 cases had missing data. None of the 198 patients with retrievable records were started on treatment. Of these 198 patients, 119 (60.1%) were aged ≥80 years at the time of death, and 93 (47%) had more than one admission to hospital before death. Prior to death, of these 198 patients, 83 (41.9%) were living in an old age home (OAH), 78 (39.4%) were bed-ridden, and 121 (61.1%) had an advance care directive such as ‘do not resuscitate’ or ‘do not investigate’ stated in the case notes. Table 2 summarises the univariate and multiple logistic regression analyses of these 198 early deaths, using deaths occurring within 1 year after notification as controls. Female sex, having a malignancy other than lung cancer, living in an OAH, drug abuser, sputum TB smear negative, sputum TB culture positive, and chest X-ray (CXR) not done or not available were independent risk factors for death before TB diagnosis.

Subgroup analysis was carried out for patients that most likely died of TB. The study group included the 30 patients who died of TB before diagnosis and the 52 patients whose deaths were initially reported as ‘sepsis or pneumonia unspecified’ but later sputum TB culture was positive. The control group was all patients who died of TB after notification (Table 3). Female sex, living in an OAH, sputum TB smear negative, sputum TB culture positive, and CXR not done or not available were independent risk factors for this group.

In the present study, the proportion of patients with TB aged ≥60 years increased by 25% from 1996 to 2010. However, over the same interval, the proportion of patients who had died before completion of treatment nearly doubled (Table 1). A substantial proportion (211 of 1061; 19.9%) of the TB-related deaths were notified after death. Over 60% of these cases were aged ≥80 years and none were started on treatment, suggesting a failure to detect TB rather than just a delay in notification. Over 60% of them had an advance care directive against resuscitation or investigation, likely indicating a concurrent terminal illness. Independent factors associated with TB notified after death were female sex, malignancies other than lung cancer, living in an OAH, drug abuse, sputum TB smear negative, sputum TB culture positive, and CXR not done. Although the recorded cause of death was TB in only 30 (14%) cases, in 52 (25%) cases the recorded cause of death was respiratory disease (predominantly pneumonia unspecified), particularly among those aged ≥80 years (19% vs 39%; P<0.005). In these cases, pulmonary TB is likely to have been the main or precipitating cause.

In the present study, the fatality rate in the first year of TB notification was 17.5% (4.1% died before TB notification and 13.4% died within 1 year after TB notification). This is much higher than rates reported earlier in Europe (7.8%9) and England and Wales (8.4%10), but similar to rates reported more recently in Taiwan (16.5%11). This is probably a reflection of differences among patient profiles in these regions, especially age and the associated co-morbidities. In the present study, 47% of patients with TB were aged ≥60 years (Table 1), whereas in the studies in Europe and the United Kingdom only 24.3%9 and 17.9%10 of the patients with TB were cohort aged ≥60 years.

Our finding that 4.1% of TB cases were notified after death is similar to rates reported in Taiwan in 2006 (4.0%12) and in the US in the 1980s (5.1%13 and 3.9%14). In all of these reports, advanced age was a consistent observation for this extreme form of delayed diagnosis. As expected from the relatively short turnover time for sputum TB smear tests and CXRs, sputum TB smear negative, and CXR unknown or not done were important risk factors for TB notified after death. The strong association between these cases and positive sputum TB culture might be explained by the fact that the sputum TB culture was the primary method of TB diagnosis, unless a diagnosis had already been made during autopsy.

Our findings that drug abusers have a higher chance of TB notification after death is in line with an earlier study that suggested such patients have difficulty completing medical evaluations.15 Drug abusers might be less aware of their TB symptoms because of the effects of the drugs taken, such as opiate suppression of the cough reflex.

Female sex was also an independent factor in the current study, similar to a previous study in Taiwan.11 This is expected, because there is a higher proportion of women among the geriatric population16 and among residents of OAH17 owing to their longer life expectancy and because conservative treatment is more frequently selected by these elderly female patients or their guardians. Patients with terminal conditions might have an advance care directive against resuscitation or investigation. An incorrect provisional diagnosis might also result from the readiness to accept a diagnosis of advanced disseminated malignancy in a patient with such an advance care directive. As lung cancer patients usually had CXR and sputum samples taken in their initial diagnostic investigation, coexisting TB could be discovered early. In addition, most lung cancer patients were diagnosed at an advanced stage and usually died within the first year after presentation.18

In our study, TB-related death occurred shortly before or after TB treatment was started, in line with findings from studies in Taiwan,19 the US,20 and Russia21 reporting a median time of 3 to 7 weeks from diagnosis or notification of TB to death. A study in Canada showed that a delay in TB treatment increased risk of death (aOR=3.3; 95% confidence interval=1.7-6.2) and intensive care unit admission (aOR=16.8; 95% confidence interval=2-144).22 Another study of hospitalised patients with TB also showed that late TB treatment guided by conventional TB culture was associated with a higher mortality than for treatment guided by polymerase chain reaction (PCR), liquid culture, positive histological findings or typical clinico-radiological manifestation.23 In settings with a high human immunodeficiency virus prevalence, the WHO advocates early empirical TB treatment based on clinical and radiological criteria in patients strongly suspected as having TB but with sputum TB smear negative, because this can improve survival.24 25

Although a timely diagnosis might not avert most non–TB-related deaths, early treatment could reduce the institutional transmission risk, because 42% of patients with TB were living in OAHs in the current study. The prevalence of active TB in OAHs has been estimated to be as high as 669 per 100 000 person-years in Hong Kong.26 The majority of patients in the present study did not have a positive sputum TB smear; however, a representative sputum sample might have been difficult to obtain from patients living in OAHs. That 73% of these patients had a positive sputum TB culture suggests that there was a sufficient degree of suspicion, either clinical or radiological, for initiation of bacteriological sampling. In total, 54 out of 211 patients who died before TB notification were recorded to have ‘pneumonia unspecified’ or ‘respiratory disease’ as the main cause of death. Past studies have shown that negative TB smear contributed to around 17% of TB transmission in San Francisco27 and Vancouver28 and even 30% in China.29 Thus, rapid diagnosis with effective isolation and early treatment can reduce transmission and even mortality. Sputum induction30 or gastric aspiration31 would improve specimen collection. However, in view of the infection risk, these bio-aerosol generating procedures would preferentially be performed in a negative pressure room with effective personal protective equipment as stipulated by the Institutional Infection Control Guidelines. Real-time PCR diagnostic tests such as Xpert® MTB/RIF assay32 may also be valuable, either as a primary diagnostic test or as an add-on test in patients previously found to be TB smear negative, to avoid the long turnover time for bacteriological cultures. In a study in Hong Kong,33 Xpert® MTB/ RIF assay was found to be a highly cost-effective strategy for TB diagnosis in terms of quality-adjusted life-years gained and lower first year mortality rate.

Higher mortality among patients with TB aged ≥80 years is a consistent finding among different TB programmes.34 The present study also found frequently missed diagnosis of TB and excessive mortality among patients aged ≥80 years who were frequently institutionalised and had multiple co-morbidities. A high index of suspicion and rapid diagnostic tools are necessary to reduce both mortality and transmission risk in a rapidly ageing population, in order to meet the WHO End TB 2035 target of a 95% reduction in TB mortality rate compared with the 2015 rate.35

This study shares an important limitation with other retrospective studies. The clinical data in this cohort were constructed from a database of the pre-assembled ‘TB programme record form’ which was not specifically designed for this study. Therefore, not all pertinent risk factors were identified and recorded. As this is a population-wide database, many health care professionals were involved and the measurement of risk factors and outcomes is less accurate and less consistent than a prospective study. Nonetheless, data from the TB programme record form have been used in previous studies on patients with TB6 and elderly patients with TB7 and were included for comparison in this study.

This study was a collaborative effort between the Hospital Authority and the Department of Health, and a database was compiled for all patients with TB treated in the public or the private sector. This study provides insight into the mortality of patients with TB and the risk factors associated with a delay in TB diagnosis. These factors include novel patient factors such as female sex, living in OAHs, advance care directives refusing further investigation or resuscitation, and drug abuse. Additional factors include lack of a representative sputum sample. which could be mitigated by sputum induction or gastric aspiration, and the relative insensitivity of sputum TB smear and long turnover time for conventional TB culture, which could be mitigated by using of real-time PCR tests. Information generated by this study will help frontline clinicians to be better aware of this important infectious disease among elderly people. Hopefully, more resources will be allocated to promote rapid diagnosis of TB for patients in high-risk scenarios in Hong Kong.

The authors would likely to thank the Nursing and General Grade staff in Department of Health and Hospital Authority for their assistance in collection and compilation of the demographical, clinical and laboratory data for this study.

Concept or design: ECC Leung, CC Leung, CK Chan, KC Chang.
Acquisition of data: TYW Mok, KS Chan, KS Lau, CH Chau, WKS Yee, WM Leung, KF Au.
Analysis or interpretation of data: WS Law, SN Lee, LB Tai.
Drafting of the article: ECC Leung, CC Leung, WM Leung, WS Law.
Critical revision for important intellectual content: All authors.

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

The authors have no conflicts of interest to disclose.

This study was approved by the Ethics Committee of the Department of Health and Ethics Committees of all hospital clusters from the Hospital Authority.

1. Demographic Statistics Section, Census and Statistics Department, Hong Kong SAR Government. Demographic Trends in Hong Kong 1981-2011. Available from: https://www.statistics.gov.hk/pub/B1120017032012XXXXB0100.pdf Accessed 13 Jul 2018. Crossref

2. Tuberculosis and Chest Service, Department of Health, Hong Kong SAR Government. Annual Report 1981. Hong Kong: Department of Health, Hong Kong SAR Government; 1981. Crossref

3. Tuberculosis and Chest Service, Department of Health, Hong Kong SAR Government. Annual Report 2011. Available from: http://www.info.gov.hk/tb_chest/doc/AnnualReport2011.pdf. Accessed 14 Jul 2018. Crossref

4. Leung CC, Yew WW, Chan CK, et al. Tuberculosis in older people: a retrospective and comparative study from Hong Kong. J Am Geriatr Soc 2002;50:1219-26. Crossref

5. World Health Organization. Definitions and reporting framework for tuberculosis—2013 revision. Geneva, Switzerland: World Health Organization; 2013. Crossref

6. Tam CM, Leung CC, Noertjojo K, Chan SL, Chan-Yeung M. Tuberculosis in Hong Kong-patient characteristics and treatment outcome. Hong Kong Med J 2003;9:83-90.

7. Chan-Yeung M, Noertjojo K, Tan J, Chan SL, Tam CM. Tuberculosis in the elderly in Hong Kong. Int J Tuberc Lung Dis 2002;6:771-9.

8. Centre for Health Protection, Department of Health, Hong Kong SAR Government. Notification & death rate of tuberculosis (all forms), 1947-2017. Available from: https://www.chp.gov.hk/en/statistics/data/10/26/43/88.html. Accessed 20 Jul 2018. Crossref

9. Lefebvre N, Falzon D. Risk factors for death among tuberculosis cases: analysis of European surveillance data. Eur Respir J 2008;31:1256-60. Crossref

10. Crofts JP, Pebody R, Grant A, Watson JM, Abubakar I. Estimating tuberculosis case mortality in England and Wales, 2001-2002. Int J Tuberc Lung Dis 2008;12:308-13.

11. Wu YC, Lo HY, Yang SL, Chu DC, Chou P. Comparing the factors correlated with tuberculosis-specific and non-tuberculosis-specific deaths in different age groups among tuberculosis-related deaths in Taiwan. PLoS ONE 2015;10:e0118929. Crossref

12. Wu YC, Lo HY, Yang SL, Chou P. Factors correlated with tuberculosis reported after death. Int J Tuberc Lung Dis 2014;18:1485-90. Crossref

13. Rieder HL, Kelly GD, Bloch AB, Cauthen GM, Snider DE Jr. Tuberculosis diagnosed at death in the United States. Chest 1991;100:678-81. Crossref

14. DeRiemer K, Rudoy I, Schecter GF, Hopewell PC, Daley CL. The epidemiology of tuberculosis diagnosed after death in San Francisco, 1986-1995. Int J Tuberc Lung Dis 1999;3:488-93.

15. Deiss RG, Rodwell TC, Garfein RS. Tuberculosis and illicit drug use: review and update. Clin Infect Dis 2009;48:72-82. Crossref

16. Census and Statistics Department, Hong Kong SAR Government. Population by Age Group and Sex. Hong Kong Population By-Census Main Report. 2015. Available from: http://www.censtatd.gov.hk/hong_kong_statistics/statistical_tables/index.jsp?charsetID=1<tableID=002. Accessed 16 Mar 2016.

17. Luk JK, Chan FH, Pau MM, Yu C. Outreach geriatric service to private old age homes in Hong Kong West Clusters. J HK Geriatr Soc 2002;11:5-11.

18. Cancer Research UK. Lung cancer survival statistics. Available from: http://www.cancerresearchuk.org/healthprofessional/cancer-statistics/statistics-by-cancer-type/lung-cancer/survival. Accessed 15 Mar 2016. Crossref

19. Lin CH, Lin CJ, Kuo YW, et al. Tuberculosis mortality: patient characteristics and causes. BMC Infect Dis 2014;14:5. Crossref

20. Oursler KK, Moore RD, Bishai WR, Harrington SM, Pope DS, Chaisson RE. Survival of patients with pulmonary tuberculosis: clinical and molecular epidemiologic factors. Clin Infect Dis 2002;34:752-9. Crossref

21. Mathew TA, Ovsyanikova TN, Shin SS, et al. Causes of death during tuberculosis treatment in Tomsk Oblast Russia. Int J Tuberc Lung Dis 2006;10:857-63.

22. Greenaway C, Menzies D, Fanning A, et al. Delay in diagnosis among hospitalized patients with active tuberculosis—predictors and outcomes. Am J Respir Crit Care Med 2002;165:927-33. Crossref

23. Lui G, Wong RY, Li F, et al. High mortality in adults hospitalized for active tuberculosis in a low HIV prevalence setting. PLoS One 2014;9:e92077. Crossref

24. Holtz TH, Kabera G, Mthiyane T, et al. Use of a WHO-recommended algorithm to reduce mortality in seriously ill patients with HIV infection and smear-negative pulmonary tuberculosis in South Africa: an observational cohort study. Lancet Infect Dis 2011;11:533-40. Crossref

25. Katagira W, Walter ND, Den Boon S, et al. Empiric TB treatment of severely ill patients with HIV and presumed pulmonary TB improves survival. J Acquir Immune Defic Syndr 2016;72:297-303. Crossref

26. Chan-Yeung M, Chan FH, Cheung AH, et al. Prevalence of tuberculous infection and active tuberculosis in old age homes in Hong Kong. J Am Geriatr Soc 2006;54:1334-40. Crossref

27. Behr MA, Warren SA, Salamon H, et al. Transmission of Mycobacterium tuberculosis from patients smear-negative for acid-fast bacilli. Lancet 1999;353:444-9. Crossref

28. Hernández-Garduño E, Cook V, Kunimoto D, Elwood RK, Black WA, FitzGerald JM. Transmission of tuberculosis from smear negative patients: a molecular epidemiology study. Thorax 2004;59:286-90. Crossref

29. Yang C, Shen X, Peng Y, et al. Transmission of Mycobacterium tuberculosis in China; a population-based molecular epidemiologic study. Clin Infect Dis 2015;61:219-27. Crossref

30. Chang KC, Leung CC, Yew WW, Tam CM. Supervised and induced sputum among patients with smear-negative pulmonary tuberculosis. Eur Respir J 2008;31:1085-90. Crossref

31. Brown M, Varia H, Bassett P, Davidson RN, Wall R, Pasvol G. Prospective study of sputum induction, gastric washing, and bronchoalveolar lavage for the diagnosis of pulmonary tuberculosis in patients who are unable to expectorate. Clin Infect Dis 2007;44:1415-20. Crossref

32. Steingart KR, Schiller I, Horne DJ, Pai M, Boehme CC, Dendukuri N. Xpert® MTB/RIF assay for pulmonary tuberculosis and rifampicin resistance in adults. Cochrane Database Syst Rev 2014;(1):CD009593. Crossref

33. You JH, Lui G, Kam KM, Lee NL. Cost-effectiveness analysis of the Xpert MTB/RIF assay for rapid diagnosis of suspected tuberculosis in an intermediate burden area. J Infect 2015;70:409-14. Crossref

34. Waitt CJ, Squire SB. A systematic review of risk factors for death in adults during and after tuberculosis treatment. Int J Tuberc Lung Dis 2011;15:871-85. Crossref

35. World Health Organization. Implementing the end TB strategy: the essentials. Available from: http://www.who.int/tb/publications/2015/The_Essentials_to_End_TB/en/.Accessed 3 May 2017. Crossref

Cardiac genetics is evolving rapidly and many new insights have recently been achieved. Genetic causes are found in various potentially lethal channelopathies and cardiomyopathies including long and short QT syndrome (LQTS and SQTS), Brugada syndrome, catecholaminergic polymorphic ventricular tachycardia (CPVT), hypertrophic cardiomyopathy (HCM), dilated cardiomyopathy (DCM), arrhythmogenic right ventricular dysplasia/cardiomyopathy (ARVD/C), Barth syndrome, and left ventricular non-compaction.1 Knowledge of genetics deepens the understanding of pathophysiology and remarkably changes the diagnosis, treatment, and genetic counselling for recurrence risk and family planning. This group is highly genetically heterogeneous (Table 12).

The genetic basis of inherited cardiac diseases in the Hong Kong Chinese population is largely unknown. The Princess Margaret Hospital provides a comprehensive cardiac genetic service. We conducted this study to review the clinical and genetic findings of 28 unrelated positive cases encountered between January 2006 and December 2015.

Diagnosis of the cardiac conditions was based on clinical assessments by a cardiologist and practice guidelines.3 4 5 The patients were referred by cardiologists from various public hospitals for genetic analysis. Only patients with positive genetic findings are reported in this study. There were seven patients with LQTS, two with Brugada syndrome, two with CPVT, nine with HCM, four with DCM, and four with ARVD/C. Local ethics board approval was obtained. Peripheral blood samples were collected from the proband after informed consent was obtained. Genomic DNA was extracted using a QIAamp Blood Kit (Qiagen, Hilden, Germany). The coding exons and the flanking introns (10 bp) of each gene were amplified by polymerase chain reaction. The primer sequences and protocol are available on request. Sanger sequencing was performed in the following order and stopped once a positive finding was detected: KCNQ1, KCNH2, KCNE1, KCNE2, and SCN5A for LQTS; SCN5A for Brugada syndrome; RYR2 for CPVT; MYH7 and MYBPC3 for HCM; LMNA for DCM; and PKP2 and DSP for ARVD/C. The order was based on prevalence according to the literature and local experience. All coding exons were amplified for each gene except selected exons 3, 8, 14, 45, 46, 47, 49, 88, 89, 90, 93, 96, 97, 100, 101, and 103 for RYR2.6 The GenBank accession numbers are shown in Table 2. The pathogenicity of novel missense variants was analysed by Alamut Visual (Interactive Biosoftware, Rouen, France) with Polymorphism Phenotyping v2 (PolyPhen-2), Sorting Intolerant from Tolerant (SIFT), MutationTaster, and Assessing Pathogenicity Probability in Arrhythmia by Integrating Statistical Evidence (APPRAISE, https://cardiodb.org/APPRAISE/) and that of novel splicing variants by Splice Site Finder-like, MaxEntScan, NNSPLIC, GeneSplicer, and Human Splicing Finder, wherever appropriate. Splicing variants were considered to be damaging if there was a >10% lower score when compared with the wild-type prediction. Allele frequencies among populations were referred to the Exome Aggregation Consortium (ExAC; http://exac.broadinstitute.org/).

During the 10-year study period more than 90 patients with channelopathies or cardiomyopathies were referred for genetic analysis. Among them, 28 unrelated patients had positive genetic results, comprising 17 males and 11 females. Their mean age at diagnosis was 39 years (range, 1-80 years). The major clinical presentations included syncope, palpitations, and abnormal electrocardiography (ECG) findings. Four patients were asymptomatic and were diagnosed following an incidental abnormal finding related to other medical issues. A family history was present in only 13 (46%) patients. All detected mutations were heterozygous, and 26 different heterozygous mutations were detected. These encompassed 11 missense, two nonsense, and five splicing mutations, as well as eight small insertions and deletions. There were six novel mutations—two in SCN5A (NM_198056.2:c.429del and c.2024-11T>A), two in MYBPC3 (NM_000256.3:c.906-22G>A and c.2105_2106del), and two in LMNA (NM_170707.3:c.73C>A and c.1209_1213dup) [Table 3]. All were considered pathogenic or likely pathogenic according to the Practice Guidelines for the Evaluation of Pathogenicity and the Reporting of Sequence Variants in Clinical Molecular Genetics by the Association for Clinical Genetic Science.7 Further clinical details and genotypes are shown in Table 3.

There were seven patients with LQTS, two with Brugada syndrome, two with CPVT, nine with HCM, four with DCM, and four with ARVD/C. Three patients with LQTS had mutations in KCNQ1 (cases 1-3) and four had mutations in KCNH2 (cases 4-7). Two patients (cases 8 and 9) with Brugada syndrome had mutations in SCN5A, including two novel mutations. Two patients (cases 10 and 11) with CPVT had mutations in RYR2. Four patients with HCM (cases 12-15) had MYH7 mutations and five (cases 16-20) had MYBPC3 mutations, including two novel mutations. Four patients with DCM (cases 21-24) had LMNA mutations, including two novel mutations. Finally, three patients with ARVD/C had PKP2 mutations (cases 25-27) and one had a DSP mutation (case 28).

This is the first report of a cardiac genetic case series among Hong Kong Chinese patients with channelopathies and cardiomyopathies. A total of 28 patients are reported, and 26 different mutations and six novel mutations have been identified. Wide genetic diversity is observed, with no common mutation found. Hereditary channelopathies and cardiomyopathies are mainly inherited in an autosomal dominant manner. Mutations can be either inherited or de novo. Risk to proband sibling(s) and first-degree relatives depends on the genetic status of the parents. Offspring of the proband have a 50% risk of inheriting the mutation. Siblings of the proband have the same risk if the mutation is transmitted from either parent. Patients carrying a mutation of these sudden arrhythmia death syndromes show incomplete penetrance. In general, a mutation carrier will show symptoms/signs in 80% of those with CPVT, 20% to 50% of ARVD/C patients, 18% to 63% of LQTS patients, 80% to 94% of SQTS patients, and 80% of patients with Brugada syndrome who have abnormal ECG findings when challenged with a sodium channel blocker.8 No exact figure is available for HCM. The data could be more specific if a particular mutation was considered alongside clinical findings and family history. Pre-symptomatic testing of at-risk family members cannot be used to predict age of onset, severity, type of symptoms, or rate of progression. Detailed clinical, ECG, and genetic characterisation of affected and unaffected family members is helpful.

Long QT syndrome is genetically heterogeneous, with at least 12 genes involved. Mutations in the four genes, KCNQ1, KCNH2, KCNE1, and KCNE2, are detected in 46%, 38%, 2%, and 1% of affected patients, respectively.8 A small proportion of patients (3%) have double heterozygous mutations in more than one disease loci.9 Specific arrhythmogenic triggers are associated with a particular subtype, such as exertion, swimming, and near-drowning for LQT1; auditory triggers and cardiac events occurring in the postpartum period for LQT2; and cardiac events during sleep or at rest for LQT3. Three patients had KCNQ1 mutations. Case 1 had recurrent syncope induced by exercise and swimming, but genetic testing confirmed LQTS type 1. Other patients had no specific provoking factor. LQTS type 2 caused by KCNH2 mutations accounts for about 38% of all LQTS.8 Four patients (cases 4-7) carried KCNH2 mutations and two (cases 4 and 6) presented with Torsades de pointes and one (case 7) had survived cardiac arrest requiring an implantable cardioverter defibrillator. Case 6 was the youngest patient, presenting at age 1 year. Genotype-guided treatment in LQTS is recommended and LQT1 responds best to beta-blockers.10 11

Brugada syndrome is characterised by cardiac conduction abnormalities (ST-segment abnormalities in leads V1-V3 on ECG and a high risk for ventricular arrhythmias) that can result in sudden death. The Shanghai Score System has been recently published for the diagnosis of Brugada syndrome.12 13 The prevalence of Brugada syndrome or its characteristic ECG pattern is reportedly higher among Asians, such as Japanese (0.14%-1.22%).14 15 16 17

Brugada syndrome is genetically heterogeneous and can be attributed to defects in at least 23 genes at the time of reporting.8 Mutations in SCN5A are detected in 11% to 14% of affected individuals in Japan and <10% in Taiwan where mutations in CACNA1C account for 1% to 7%.18 Approximately 65% to 70% of patients remain genetically undiagnosed. Expressivity is variable and penetrance is incomplete and low.

Conventionally, Brugada syndrome has been described as a monogenic disease that has autosomal dominant inheritance with incomplete penetrance; it is caused by rare genetic variants with a large effect size. Most individuals diagnosed with Brugada syndrome have an affected parent. The proportion of cases caused by a de-novo mutation is approximately 1%. Recent studies indicate that genetic inheritance is likely more complex, and models of an oligogenic disorder or susceptibility risk/genetic predisposition have been suggested.19 20 21 22

Among the two patients in this series, none had a positive family history. Symptoms were more non-specific, such as palpitation and syncope. It is noteworthy that convulsion can be a presentation of channelopathies (case 8). Clinical suspicion should be higher with more specific investigations, such as exercise-stress ECG and flecainide challenge tests, are required in order to reveal the real culprit. Sudden cardiac death can be the first presenting symptom in Brugada syndrome.

Two novel mutations are described in SCN5A: c.429del and c.2024-11T>A. The former is predicted to cause a frameshift and premature protein truncation. The latter is predicted to abolish the acceptor splice site and create a cryptic site upstream. At the time of reporting, both are absent from controls in the Exome Sequencing Project, 1000 Genomes Project, and ExAC. SCN5A mutations can cause either LQTS or Brugada syndrome.

Catecholaminergic polymorphic ventricular tachycardia can present with syncope and sudden death during physical exertion or emotion, due to catecholamine-induced bidirectional ventricular tachycardia, polymorphic ventricular tachycardia or ventricular fibrillation. The reported mean age of onset is between 7 and 12 years.8 Exercise stress testing or an adrenaline provocation test may induce ventricular arrhythmia and enable a clinical diagnosis. About half of these cases are related to a dominantly inherited RYR2 gene mutation, with a small proportion (1%-2%) related to recessively inherited CASQ2 gene mutations. RYR2 is a large gene with 105 exons. Tier testing has been proposed by Medeiros-Domingo et al.6 First-tier RYR2 genetic testing of the 16 selected exons allows identification of about 65% of CPVT cases. There were two paediatric CPVT patients (cases 10 and 11) in our series, with two known disease-causing mutations detected, namely NM_001035.2(RYR2):c.11836G>A (p.Gly3946Ser)23 24 25 26 and c.14848G>A (p.Glu4950Lys).23 24 Both mutations were detected in first-tier screening.

Hypertrophic cardiomyopathy is the most prevalent hereditary cardiac disease, causing about one third of sudden cardiac deaths in young athletes. Its prevalence in China is approximately 1 in 1250.27 The clinical manifestations are markedly variable, ranging from asymptomatic to sudden cardiac death. Genetic testing provides an accurate diagnosis in the probands and enables screening of asymptomatic family members. Although the genetic background of HCM is heterogeneous, involving at least 30 genes, MYH7 and MYBPC3 are the most common and each accounts for approximately 40%.8

Nine patients with HCM are reported here: four had known MYH7 mutations and five had MYBPC3 mutations, including two novel mutations. NM_000256.3:c.906-22G>A was detected in case 16 and was a novel variant. Neither population frequency nor known pathogenicity have been reported. In-silico analysis showed creation of a novel acceptor site and insertion of 20 nucleotides into exon 10. This conceivably would lead to a frameshift and premature protein termination. Exon 10 of MYBPC3 is a microexon in which the stability of its original splicing site is easily disrupted by intronic variants. A similar mutation has been reported as c.906-36G>A.28 Nonetheless, cDNA analysis was not performed. NM_000256.3(MYBPC3):c.1223+1G>A at the critical canonical +1 splice site is also novel. In addition, other known disease-causing splicing mutations affecting the same nucleotide have been reported.26 29 30 Case 19 had two variants detected in MYBPC3 (c.2215G>A and c.3624del). The small deletion c.3624del is a mutation known to cause HCM in the Chinese population31 and predicted to cause a frameshift and premature termination of the protein. The missense variant c.2215G>A is as yet unreported and is predicted by in-silico analyses to cause an amino acid change from glutamate to lysine at codon 739 and probably damage. At the time of reporting, the variant is absent from controls in the Exome Sequencing Project, 1000 Genomes Project, and ExAC databases. This variant is considered to have uncertain significance. The mother of the patient in case 19 was available for testing. She was 48 years old at the time of genetic testing, asymptomatic, and heterozygous for c.3624del only. Hence, the two variants c.2215G>A and c.3624del of MYBPC3 were in-trans in the patient and elder brother of the patient in case 19. Both had a more severe form of HCM, with a younger onset.

Familial DCM is a group of genetically heterogeneous disorders. Laminopathy can manifest as several allelic disorders affecting muscle, nerve, adipose, and vascular tissues; one of them is cardiomyopathy, dilated 1A. We identified four patients with DCM, two of whom also had proximal muscle weakness. Two novel mutations in LMNA were detected (c.73C>A and c.1209_1213dup). NM_005572.3(LMNA):c.73C>A is a novel variant that is predicted to be deleterious by SIFT, probably causing damage according to PolyPhen-2 and disease-causing according to MutationTaster. Other missense mutations have been reported in the same amino acid codon.32 33 34 NM_005572.3(LMNA):c.1609-1G>A is predicted to significantly affect splicing by in-silico analysis. At the time of reporting, all variants are absent from controls in the Exome Sequencing Project, 1000 Genomes Project, and ExAC. In case 22 with NM_005572.3(LMNA):c.73C>A, one of the parents died of chronic heart failure in the fourth decade of life, and one sibling died of heart block and chronic heart failure with a diagnosis of muscular dystrophy at age 38 years. Nonetheless, there was no sample left for genotyping.

Arrhythmogenic right ventricular dysplasia/cardiomyopathy is associated with fibrofatty replacement of cardiomyocytes, ventricular tachyarrhythmias, and sudden cardiac death. Although the right ventricle is primarily affected in this condition, left-dominant arrhythmogenic cardiomyopathy has also been described, and mutations have been identified in DSP as well as in other genes.35 Four patients are reported here, with three having mutations in PKP2 and one in DSP. Interestingly, the patient in case 26 presented at age 80 years with episodic palpitations. His ECG results showed paroxysmal ventricular tachycardia. He had a deletion in PKP2, c.1125_1132del (p.Phe376Alafs*8), resulting in a truncated incomplete protein product. Age of onset in patients with PKP2 mutations is older than that of the patient with DSP mutation. The latter patient (case 28) died at age 23 years, with sudden collapse as the first presentation.

Primary arrhythmogenic disorders including LQTS/SQTS, CPVT, Brugada syndrome, and cardiomyopathies account for about one third of sudden cardiac deaths in the young.36 Identification of a pathogenic variant can solve the diagnostic mystery, provide relief to the family, and enable family screening and counselling for other at-risk family members. In some developed countries, molecular autopsy is an essential part of a formal forensic investigation in unexplained sudden death.37 We support the implementation of molecular autopsy in routine autopsy investigation of sudden cardiac death victims. Our group has conducted the first local prospective study to determine the prevalence and types of sudden arrhythmia death syndrome underlying sudden cardiac death among local young victims through clinical and molecular autopsy of sudden cardiac death victims and clinical and genetic evaluation of their first-degree relatives (http://www.sadshk.org/en/medical_research.php). Such data can serve as the groundwork for the feasibility of implementation of such investigations in Hong Kong.

Genetic tests for cardiac conditions can aid diagnosis and guide treatment. Nonetheless, there are limitations that complicate the translational use of genetic results in patient care, such as incomplete penetrance, variable expressivity, and findings of variants of uncertain significance. In addition, since the genetic heterogeneity is large among cardiomyopathies and channelopathies and more genes are yet to be discovered, a negative genetic finding does not necessarily exclude a genetic basis of disease in patients.

Major limitations of the current study include its small sample size, incomplete family data for co-segregation study, and lack of functional study of novel variants. We observed a lower rate of use of genetic tests in early years that might have been due to insufficient awareness among clinicians about the clinical usefulness of such tests for channelopathies and cardiomyopathies. Clinical indications published in an expert consensus statement on the state of genetic testing for channelopathies and cardiomyopathies from the Heart Rhythm Society and European Heart Rhythm Association provide a good reference to determine when a genetic test should be requested.5 In our hospital, referral information can be accessed on http://kwcpath.home/genetics/ and more information about genetic service provision in public hospitals is available in the Hong Kong Hospital Authority Genetic Test Formulary (http://gtf.home/). A comprehensive system of cardiac genetics service is required for an efficient referral system, resource funding, training, and appropriate long-term follow-up.

We present the phenotypic and genotypic characteristics of 28 unrelated Hong Kong Chinese patients diagnosed across a 10-year period. For each disease entity, it was beyond our reach in the past decade to exhaustively screen for all known genes. We therefore focus on the most common ones when investigating cardiac genetics. Even so, genetic analysis can provide an accurate diagnosis and is of utmost importance for the management of patients and their families. Non-penetrance, variable expressivity, phenotype-genotype correlation, susceptibility risk, and digenic inheritance have been reported. Genetic testing also allows for genetic counselling on the recurrence risk. Correct interpretation of genetic findings for careful genetic counselling requires professional expertise with relevant experience in both clinical medicine and molecular genetics. Next-generation sequencing will improve diagnostic performance in this genetically heterogeneous group of channelopathies and cardiomyopathies, and could become a mainstay diagnostic tool.

All authors have made substantial contributions to the concept or design of this study; acquisition of data; analysis or interpretation of data; drafting of the article; and critical revision for important intellectual content.

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

All authors have no conflicts of interest to disclose. 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.

Local ethical approval of this study was obtained (KW/EX/09-155).

1. Herman A, Bennett MT, Chakrabarti S, Krahn AD. Life threatening causes of syncope: channelopathies and cardiomyopathies. Auton Neurosci 2014;184:53-9. Crossref

2. Ackerman MJ, Marcou CA, Tester DJ. Personalized medicine: genetic diagnosis for inherited cardiomyopathies/channelopathies. Rev Esp Cardiol (Engl Ed) 2013;66:298-307. Crossref

3. Priori SG, Wilde AA, Horie M, et al. HRS/EHRA/APHRS expert consensus statement on the diagnosis and management of patients with inherited primary arrhythmia syndromes: document endorsed by HRS, EHRA, and APHRS in May 2013 and by ACCF, AHA, PACES, and AEPC in June 2013. Heart Rhythm 2013;10:1932-63. Crossref

4. Authors/Task Force members, Elliott PM, Anastasakis A, et al. 2014 ESC guidelines on diagnosis and management of hypertrophic cardiomyopathy: the task force for the diagnosis and management of hypertrophic cardiomyopathy of the European Society of Cardiology (ESC). Eur Heart J 2014;35:2733-79. Crossref

5. Ackerman MJ, Priori SG, Willems S, et al. HRS/EHRA expert consensus statement on the state of genetic testing for the channelopathies and cardiomyopathies: this document was developed as a partnership between the Heart Rhythm Society (HRS) and the European Heart Rhythm Association (EHRA). Europace 2011;13:1077-109. Crossref

6. Medeiros-Domingo A, Bhuiyan ZA, Tester DJ, et al. The RYR2-encoded ryanodine receptor/calcium release channel in patients diagnosed previously with either catecholaminergic polymorphic ventricular tachycardia or genotype negative, exercise-induced long QT syndrome: a comprehensive open reading frame mutational analysis. J Am Coll Cardiol 2009;54:2065-74. Crossref

7. Wallis Y, Payne S, McAnulty C, et al. Practice guidelines for the evaluation of pathogenicity and the reporting of sequence variants in clinical molecular genetics. Association for Clinical Genetic Science; 2013.

8. Adam MP, Ardinger HH, Pagon RA, Wallace SE, editors. GeneReviews [internet]. Seattle (WA): GeneReviews; 1993. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1116/. Accessed 19 Feb 2018.

9. Bokil NJ, Baisden JM, Radford DJ, Summers KM. Molecular genetics of long QT syndrome. Mol Genet Metab 2010;101:1-8. Crossref

10. Giudicessi JR, Ackerman MJ. Genotype- and phenotype-guided management of congenital long QT syndrome. Curr Probl Cardiol 2013;38:417-55. Crossref

11. Priori SG, Napolitano C, Schwartz PJ, et al. Association of long QT syndrome loci and cardiac events among patients treated with beta-blockers. JAMA 2004;292:1341-4. Crossref

12. Antzelevitch C, Yan GX, Ackerman MJ, et al. J-Wave syndromes expert consensus conference report: emerging concepts and gaps in knowledge. Europace 2017;19:665-94.

13. Antzelevitch C, Yan GX, Ackerman MJ, et al. J-Wave syndromes expert consensus conference report: emerging concepts and gaps in knowledge. J Arrhythm 2016;32:315-39. Crossref

14. Furuhashi M, Uno K, Tsuchihashi K, et al. Prevalence of asymptomatic ST segment elevation in right precordial leads with right bundle branch block (Brugada-type ST shift) among the general Japanese population. Heart 2001;86:161-6. Crossref

15. Matsuo K, Akahoshi M, Nakashima E, et al. The prevalence, incidence and prognostic value of the Brugada-type electrocardiogram: a population-based study of four decades. J Am Coll Cardiol 2001;38:765-70. Crossref

16. Sakabe M, Fujiki A, Tani M, Nishida K, Mizumaki K, Inoue H. Proportion and prognosis of healthy people with coved or saddle-back type ST segment elevation in the right precordial leads during 10 years follow-up. Eur Heart J 2003;24:1488-93. Crossref

17. Hiraoka M. Brugada syndrome in Japan. Circ J 2007;71 Suppl A:A61-8.

18. Juang JM, Tsai CT, Lin LY, et al. Unique clinical characteristics and SCN5A mutations in patients with Brugada syndrome in Taiwan. J Formos Med Assoc 2015;114:620-6. Crossref

19. Bezzina CR, Barc J, Mizusawa Y, et al. Common variants at SCN5A-SCN10A and HEY2 are associated with Brugada syndrome, a rare disease with high risk of sudden cardiac death. Nat Genet 2013;45:1044-9. Crossref

20. Behr ER, Savio-Galimberti E, Barc J, et al. Role of common and rare variants in SCN10A: results from the Brugada syndrome QRS locus gene discovery collaborative study. Cardiovasc Res 2015;106:520-9. Crossref

21. Gourraud JB, Barc J, Thollet A, et al. The Brugada syndrome: a rare arrhythmia disorder with complex inheritance. Front Cardiovasc Med 2016;3:9. Crossref

22. Juang JJ, Horie M. Genetics of Brugada syndrome. J Arrhythm 2016;32:418-25. Crossref

23. Priori SG, Napolitano C, Memmi M, et al. Clinical and molecular characterization of patients with catecholaminergic polymorphic ventricular tachycardia. Circulation 2002;106:69-74. Crossref

24. Jabbari J, Jabbari R, Nielsen MW, et al. New exome data question the pathogenicity of genetic variants previously associated with catecholaminergic polymorphic ventricular tachycardia. Circ Cardiovasc Genet 2013;6:481-9. Crossref

25. Kawamura M, Ohno S, Naiki N, et al. Genetic background of catecholaminergic polymorphic ventricular tachycardia in Japan. Circ J 2013;77:1705-13. Crossref

26. Xiong HY, Alipanahi B, Lee LJ, et al. RNA splicing. The human splicing code reveals new insights into the genetic determinants of disease. Science 2015;347:1254806. Crossref

27. Zou Y, Song L, Wang Z, et al. Prevalence of idiopathic hypertrophic cardiomyopathy in China: a population-based echocardiographic analysis of 8080 adults. Am J Med 2004;116:14-8. Crossref

28. Frank-Hansen R, Page SP, Syrris P, McKenna WJ, Christiansen M, Andersen PS. Micro-exons of the cardiac myosin binding protein C gene: flanking introns contain a disproportionately large number of hypertrophic cardiomyopathy mutations. Eur J Hum Genet 2008;16:1062-9. Crossref

29. Millat G, Bouvagnet P, Chevalier P, et al. Prevalence and spectrum of mutations in a cohort of 192 unrelated patients with hypertrophic cardiomyopathy. Eur J Med Genet 2010;53:261-7. Crossref

30. Waldmuller S, Muller M, Rackebrandt K, et al. Array-based resequencing assay for mutations causing hypertrophic cardiomyopathy. Clin Chem 2008;54:682-7. Crossref

31. Liu X, Jiang T, Piao C, et al. Screening mutations of MYBPC3 in 114 unrelated patients with hypertrophic cardiomyopathy by targeted capture and next-generation sequencing. Sci Rep 2015;5:11411. Crossref

32. Narula N, Favalli V, Tarantino P, et al. Quantitative expression of the mutated lamin A/C gene in patients with cardiolaminopathy. J Am Coll Cardiol 2012;60:1916-20. Crossref

33. Vytopil M, Benedetti S, Ricci E, et al. Mutation analysis of the lamin A/C gene (LMNA) among patients with different cardiomuscular phenotypes. J Med Genet 2003;40:e132. Crossref

34. Yuan WL, Huang CY, Wang JF, et al. R25G mutation in exon 1 of LMNA gene is associated with dilated cardiomyopathy and limb-girdle muscular dystrophy 1B. Chin Med J (Engl) 2009;122:2840-5.

35. Sen-Chowdhry S, Syrris P, Prasad SK, et al. Left-dominant arrhythmogenic cardiomyopathy: an under-recognized clinical entity. J Am Coll Cardiol 2008;52:2175-87. Crossref

36. Semsarian C, Hamilton RM. Key role of the molecular autopsy in sudden unexpected death. Heart Rhythm 2012;9:145-50. Crossref

37. Torkamani A, Muse ED, Spencer EG, et al. Molecular autopsy for sudden unexpected death. JAMA 2016;316:1492-4. Crossref