Preimplantation genetic diagnosis (PGD) refers to the determination of genotype of embryos before transfer during in-vitro fertilisation (IVF) cycles. The technique can prevent couples at risk for a serious genetic disease from having an affected fetus, and therefore protects couples from the psychological trauma associated with carrying an affected child, termination of pregnancy, or recurrent miscarriage. During PGD, one blastomere is biopsied from each day-3 embryo (6-8 cells) and, after whole genome amplification (WGA), mutations can be detected directly by minisequencing and/or indirectly by linkage analysis. The first PGD baby was born in 1989 following PGD for a sex-linked genetic disease using polymerase chain reaction (PCR) for sex determination.1

According to a recent report of the European Society of Human Reproduction and Embryology (ESHRE) PGD Consortium,2 there were more than 4500 PGD cycles performed for monogenetic diseases worldwide in 2002 to 2012. The Centre of Assisted Reproduction and Embryology, Queen Mary Hospital–The University of Hong Kong (HKU-QMH CARE) achieved the first live birth after PGD for alpha-thalassaemia in 2005.3 Due to the very limited amount of DNA available for diagnosis in a single cell, the major technical challenge of PGD is contamination and allele dropout, which may result in misdiagnosis. In August 2007, we adopted the platform of WGA for PGD. Such WGA amplifies a major portion of the genome of single cells with good reproducibility,4 and enables direct mutation detection and linkage analysis simultaneously for accurate determination of the genotype of embryos. We reported our first live birth after PGD for Huntington’s disease (HD) using WGA in 2009.5 We now report the outcome of over 100 PGD cycles for monogenetic diseases at the HKU-QMH CARE.

Data from all treatment cycles were stored in a database and PGD cases were coded for indication. The data of all PGD cycles for monogenetic diseases performed in the Department of Obstetrics and Gynaecology, Queen Mary Hospital/The University of Hong Kong from 1 August 2007 to 30 April 2014 were retrieved. Depending on the monogenetic disease, the definitive mutation(s) responsible for the disease were confirmed in accredited genetic laboratories, including the Department of Pathology, Queen Mary Hospital; Clinical Genetic Service, Department of Health, Hong Kong SAR; and Molecular Pathology Division, Hong Kong Sanatorium & Hospital. Preimplantation genetic diagnosis was offered to couples with a defined genetic disease, irrespective of whether the couples had a previous affected pregnancy. All couples were extensively counselled by the reproductive medicine subspecialists and a clinical geneticist on the potential risks of IVF, intracytoplasmic sperm injection (ICSI), and PGD. The couples decided whether to proceed to PGD after non-directive informative counselling. They were also advised to have a confirmatory prenatal diagnosis for the ensuing pregnancy. Depending on the availability of the test, couples could choose an invasive or non-invasive prenatal diagnostic test to confirm the diagnosis by an accredited genetic laboratory different from the PGD laboratory.

The details of the protocols for the ovarian stimulation regimen, gamete handling, and frozen-thawed embryo transfer (FET) have been previously described.6 Surplus good-quality embryos unaffected by monogenetic diseases were vitrified by the CVM Vitrification System (CryoLogic, Mulgrave, Australia). If the patient did not become pregnant in the fresh cycle, the vitrified embryos were warmed and replaced in a subsequent FET cycle.

The HKU-QMH CARE has been performing PGD for about 10 years. The procedures for PGD have been previously described.5 In brief, embryo biopsy was performed on day 3 at the 6 to 8 cells stage, with one blastomere biopsied. Whole genome amplification by multiple displacement amplification was performed on a single blastomere.5 In all the PGD cases for monogenetic diseases, apart from direct mutation detection (except for those involving large deletions), linkage analysis was performed by two to 10 microsatellite markers flanking the mutation to reduce possible errors due to allelic dropout. Aneuploidy was not determined because of lack of indications. When required, human leukocyte antigen (HLA) typing was performed by PCR-based sequence specific primer (Collaborative Transplant Study, University of Heidelberg, Heidelberg, Germany) in the same setting as for PGD.

Between 1 August 2007 and 30 April 2014, 76 couples initiated 124 cycles for monogenetic diseases. The median age of the women was 35 (range, 26-41) years. Embryo biopsy and PGD were performed in 101 cycles, including three cycles of combined HLA typing and beta-thalassaemia. The mean number of embryos biopsied per oocyte retrieval cycle was 6.1 (761/124). A total of 761 blastomeres were biopsied and a conclusive diagnosis was obtained for 692 (91%) blastomeres. An inconclusive diagnosis during PGD could be due to failure in WGA or aneuploidy. Preimplantation genetic diagnosis was cancelled in 23 (18.5%) cycles after initiation of stimulation because of poor responses (13 cycles), risk for ovarian hyperstimulation syndrome (OHSS; 4 cycles), no mature oocytes available (3 cycles), failed fertilisation (2 cycles), or an embryologist was unavailable for PGD (1 cycle). In case of poor response (<4 good-quality embryos on day 3), cleavage-stage embryos were frozen, subsequently thawed, and pooled with fresh embryos from the following stimulated cycle for PGD. When there was excessive ovarian response and a patient was at risk for OHSS, all cleavage-stage embryos were cryopreserved and PGD was performed in the hormone replacement treatment cycles.

Overall, 92 PGD cycles proceeded to embryo transfer with one or two blastocysts replaced (mean, 1.8), resulting in an ongoing pregnancy rate (pregnancy beyond 8-10 weeks of gestation) of 28.2% per initiated cycle, 38.0% per transfer, and implantation rate of 35.2% (Table 1). Embryo transfer was cancelled in nine (8.9%) cycles after PGD due to no genetically transferrable embryos available (4 cycles), no HLA-matched embryo (1 cycle), or patient at risk for OHSS (4 cycles). From August 2012 onwards, all genetically transferrable embryos were cryopreserved after PGD when patients were at risk for OHSS.

There were 16 cycles of natural-cycle FET for PGD blastocysts, resulting in an ongoing pregnancy rate of 37.5% per transfer. The mean number of blastocysts replaced in the FET cycle was 1.3. The implantation rate in the FET cycle was 30.0%. The cumulative ongoing pregnancy rate was 33.1% per initiated cycle (Table 1).

The indications for PGD are listed in Table 2. The most frequent indication for PGD was thalassaemia (70.2%), followed by spinocerebellar ataxia type 3 (SCA3; 4.8%), and HD (4.0%). Preimplantation genetic diagnosis was performed for 20 monogenetic diseases. Successful pregnancy was achieved after PGD for 14 genetic diseases. The pregnant women were referred to the maternal-fetal medicine team at Tsan Yuk Hospital for counselling and confirmation of the genetic diseases by prenatal diagnosis or postnatal cord blood genetic tests. The latter was chosen by some of the patients who worried about the risk of miscarriage associated with invasive prenatal tests. Based on the available results of the confirmation genetic tests, no misdiagnosis was found in this small series.

In 2014, the ESHRE PGD Consortium published data for 1597 PGD cycles from 60 PGD centres in Europe in 2009.7 When comparing our PGD data with those of the ESHRE PGD Consortium, we have a comparable mean number of embryos biopsied per oocyte retrieval cycle (6.1 vs 6.6) and similar mean number of embryo transfers per retrieval cycle (1.4 vs 1.3). The ESHRE PGD Consortium reported a clinical pregnancy rate of 30.2% per transfer, while the ongoing pregnancy rate per transfer in our centre was 38.0% for fresh transfer and 37.5% for FET. The ongoing pregnancy rate per transfer in the IVF-ICSI cycle without PGD was 36% in our centre. The implantation rate of PGD embryos from the ESHRE PGD Consortium was 21.3% and in our programme was 35.2% for fresh cycles and 30.0% for FET cycles. The implantation rate in our IVF-ICSI programme without PGD was 27.7%.

A limitation of the present study was that the number of cases of each genetic disease involved was small. Many genetic diseases had only one case in this cohort of patients. The usefulness of PGD in these cases needs to be confirmed in a larger cohort of patients.

The cancellation rate for PGD after initiation of stimulation was 18.5% (23/124). The major reason for cancellation was poor response leading to a small number of good-quality embryos available for PGD. In such circumstance, cleavage-stage embryos were frozen and batched for the next stimulated PGD cycle. By increasing the number of embryos tested per PGD cycle, this ‘batching’ strategy increased the chance of having disease-free embryos for transfer. The strategy also enabled patients to have the best-quality embryo chosen, instead of experiencing multiple cancellations of embryo transfer after PGD.

For those cases proceeding to PGD, 8.9% resulted in no embryo transfer. Embryo transfer was cancelled because of the risk for OHSS or no genetically transferable embryos available. In cases at risk for OHSS, good-quality blastocysts were vitrified for subsequent FET cycles. Some reports have suggested that FET cycles may result in higher pregnancy and implantation rates than stimulated cycles8 9 10 due to the better receptivity of the endometrium without gonadotropin stimulation. An additional benefit of FET lies in being free of risk for OHSS when compared with fresh embryo transfer.

Among the 124 cycles initiated for PGD during the study period, alpha- and beta-thalassaemia (autosomal recessive disorders) accounted for 70.2% of the PGD cases. This is due to a high percentage of carriers in the population in Hong Kong, with a prevalence of 4.5% and 2.8% for alpha-thalassaemia (South-East Asian deletion type) and beta-thalassaemia, respectively.11 12 The second most frequent indication for PGD was neurodegenerative disorders such as SCA3 and HD, which are inherited in an autosomal dominant fashion. These conditions accounted for 8.9% of PGD cycles. The local prevalence of HD is estimated to be four per million population.13 14 There were 16 SCA cases diagnosed in three hospitals in Hong Kong within 3 years, and SCA3 accounted for 75% of the cases.15

It is noteworthy that 4.8% of the PGD cases were for cancer predisposition. However, it is always controversial to offer PGD to couples with inherited mutations of late-onset reduced penetrance cancer predisposition such as breast cancer.16 Additional controversy lies in the fact that PGD does not remove all the risks associated with the disease. Other known risk factors involved in breast cancer include obesity, use of hormone therapies (progestin and oestrogen), increased breast tissue density, alcohol use, and physical inactivity.17 In 2003, the ESHRE Ethics Task Force published a recommended multidisciplinary approach to the application of PGD, stating that PGD for multifactorial diseases such as BRCA mutation is acceptable notwithstanding the uncertainties about the genetic predisposition and the epigenetic influence.18 The United Kingdom Human Fertilisation and Embryology Authority also accepted PGD for various hereditary cancer syndromes, including familial adenomatous polyposis, neurofibromatosis type 1 and type 2, and von Hippel–Lindau syndrome.

Apart from the monogenetic diseases, other genetic tests such as HLA typing can be done on the same samples. In these cases, the embryos were not treated differently from those without HLA typing. Therefore, it is unlikely that the additional test would affect the ongoing pregnancy rate or implantation rate. Without indication, aneuploidy screening was not performed at the same time as PGD for monogenetic diseases. In our series, there were two pregnancies complicated by aneuploidy (trisomy 21 and trisomy 13 for each). Both patients were young (age, 31 years) with no previous pregnancy affected by aneuploidy. The usefulness of preimplantation aneuploidy screening in young patients is still under debate.

In our centre, we always consider each case individually and take into account the family histories of the couples, especially for late-onset genetic diseases such as SCA3 and HD and those with reduced penetrance and multifactorial cancer predisposition such as breast cancer. Patients were extensively counselled by a geneticist before referral for PGD treatment. Informative and non-directive counselling by specialists in reproductive medicine was also given to patients. Options other than PGD—such as prenatal diagnosis after becoming pregnant without PGD, gamete donation, embryo donation, adoption, or remaining childless—were discussed. Patients who became pregnant after PGD were referred to the prenatal diagnosis centre in Tsan Yuk Hospital for counselling. Prenatal diagnosis was encouraged for confirmation of the genetic status of the fetus. If patients refused to undergo invasive prenatal testing, postnatal cord blood genetic confirmation may be considered if the offspring will benefit from the early surveillance or postnatal treatment. When encountering controversial PGD cases, meetings were held to discuss the case with geneticists, obstetricians, specialists in prenatal diagnosis, ethicists, and other relevant specialists such as oncologists before offering PGD.

Preimplantation genetic diagnosis is a reliable method with ongoing pregnancy rate and implantation rate similar to those with IVF and ICSI. When couples have known genetic diseases, they should be counselled before pregnancy for preimplantation genetic diseases as an alternative to prenatal diagnosis, even when they do not have a previous affected pregnancy.

We would like to thank the patients, nurses, clinicians, technicians, and embryologists at the Centre of Assisted Reproduction and Embryology, Queen Mary Hospital–The University of Hong Kong for their contribution in the PGD programme.

The authors declare that they have no conflict of interest.

1. Handyside AH, Kontogianni EH, Hardy K, Winston RM. Pregnancies from biopsied human preimplantation embryos sexed by Y-specific DNA amplification. Nature 1990;344:768-70. Crossref

2. Harper JC, Wilton L, Traeger-Synodinos J, et al. The ESHRE PGD Consortium: 10 years of data collection. Hum Reprod Update 2012;18:234-47. Crossref

3. Chan V, Ng EH, Yam I, Yeung WS, Ho PC, Chan TK. Experience in preimplantation genetic diagnosis for exclusion of homozygous alpha degrees thalassemia. Prenat Diagn 2006;26:1029-36. Crossref

4. Hellani A, Coskun S, Benkhalifa M, et al. Multiple displacement amplification on single cell and possible PGD applications. Mol Hum Reprod 2004;10:847-52. Crossref

5. Chow JF, Yeung WS, Lau EY, et al. Singleton birth after preimplantation genetic diagnosis for Huntington disease using whole genome amplification. Fertil Steril 2009;92:828.e7-10. Crossref

6. Ng EH, Yeung WS, Lau EY, So WW, Ho PC. High serum oestradiol concentrations in fresh IVF cycles do not impair implantation and pregnancy rates in subsequent frozen-thawed embryo transfer cycles. Hum Reprod 2000;15:250-5. Crossref

7. Moutou C, Goossens V, Coonen E, et al. ESHRE PGD Consortium data collection XII: cycles from January to December 2009 with pregnancy follow-up to October 2010. Hum Reprod 2014;29:880-903. Crossref

8. Evans J, Hannan NJ, Edgell TA, et al. Fresh versus frozen embryo transfer: backing clinical decisions with scientific and clinical evidence. Hum Reprod Update 2014;20:808-21. Crossref

9. Roque M, Lattes K, Serra S, et al. Fresh embryo transfer versus frozen embryo transfer in in vitro fertilization cycles: a systematic review and meta-analysis. Fertil Steril 2013;99:156-62. Crossref

10. Shapiro BS, Daneshmand ST, Garner FC, Aguirre M, Hudson C. Clinical rationale for cryopreservation of entire embryo cohorts in lieu of fresh transfer. Fertil Steril 2014;102:3-9. Crossref

11. Lau YL, Chan LC, Chan YY, et al. Prevalence and genotypes of alpha- and beta-thalassemia carriers in Hong Kong—implications for population screening. N Engl J Med 1997;336:1298-301. Crossref

12. Sin SY, Ghosh A, Tang LC, Chan V. Ten years’ experience of antenatal mean corpuscular volume screening and prenatal diagnosis for thalassaemias in Hong Kong. J Obstet Gynaecol Res 2000;26:203-8. Crossref

13. Leung CM, Chan YW, Chang CM, Yu YL, Chen CN. Huntington’s disease in Chinese: a hypothesis of its origin. J Neurol Neurosurg Psychiatry 1992;55:681-4. Crossref

14. Chang CM, Yu YL, Fong KY, et al. Huntington’s disease in Hong Kong Chinese: epidemiology and clinical picture. Clin Exp Neurol 1994;31:43-51.

15. Lau KK, Lam K, Shiu KL, et al. Clinical features of hereditary spinocerebellar ataxia diagnosed by molecular genetic analysis. Hong Kong Med J 2004;10:255-9.

16. Konstantopoulou I, Pertesi M, Fostira F, Grivas A, Yannoukakos D. Hereditary cancer predisposition syndromes and preimplantation genetic diagnosis: where are we now? J BUON 2009;14 Suppl 1:S187-92.

17. Emens LA, Jaffee EM. Leveraging the activity of tumor vaccines with cytotoxic chemotherapy. Cancer Res 2005;65:8059-64. Crossref

18. Shenfield F, Pennings G, Devroey P, Sureau C, Tarlatzis B, Cohen J; ESHRE Ethics Task Force. Taskforce 5: preimplantation genetic diagnosis. Hum Reprod 2003;18:649-51. Crossref

‘Revolving door syndrome’ was a phrase coined by Gordon1 in 1995 to describe the problem of recurrent return of older people to hospital shortly after discharge. Readmission is common among medical patients, especially the elderly population, and is a poor outcome for the health care system.2 3 A retrospective analysis in 2007 showed the overall 30-day unplanned readmission rate of medical patients was 16.7%.4 A study in Hong Kong West Cluster (HKWC) showed the 28-day readmission rate for elderly patients discharged from a geriatric convalescent hospital to be 21.6%.5

To date, different programmes and strategies have been described to reduce hospital readmission. These include comprehensive geriatric assessment, discharge planning, adopting a case manager approach, post-discharge support services, early intervention for ad-hoc medical problems, and use of telephone nursing services.6 7 8 9 10 11 It has been advocated that in order to achieve better efficacy, any programme that aims to prevent hospital readmission should focus on patients at high risk.12 In Hong Kong, the Hospital Authority (HA) has developed a validated prediction model named “Hospital Admission Risk Reduction Program for the Elderly” (HARRPE) to identify older people at high risk of readmissions.13 The HARRPE score comprises 14 predictors that are categorised into socio-demographic data, prior utilisation of accident and emergency department (AED) and medical ward admission in the past 1 year, co-morbidity, and current index admission. The higher the HARRPE score (ranges from 0 to 1), the higher the readmission risk.

In Hong Kong, a pilot Integrated Discharge Support Program for Elderly Patients (IDSP) was launched in three hospitals, namely the United Christian Hospital in 2008, Princess Margaret Hospital in 2008, and Tuen Mun Hospital in 2009. The programme targeted patients aged ≥60 years admitted to these hospitals with a HARRPE score of ≥0.2. It aimed to reduce the risk of AED attendance and hospital readmission through better discharge planning and post-discharge support. Preliminary results showed it successfully reduced AED attendance, emergency admission, and hospital bed days.14

In view of the positive results of IDSP and based on the recommendation of the Elderly Commission, the Financial Secretary of the Hong Kong SAR announced in the 2011/2012 budgets that the Government would allocate additional recurrent funding to make it a regular service to all districts. In addition, a new case management approach, Integrated Care Model (ICM), was added. Such new programme has been renamed Integrated Care and Discharge Support for elderly patients (ICDS).

In January 2012, HKWC launched the ICDS and involves hospital and community components (Fig). For the hospital component, risk stratification, comprehensive geriatric assessment, and discharge planning are performed. Link nurses (who serve as ‘link’ between in-patients and community services) work with geriatricians to perform multidimensional assessments for home-dwelling older patients aged ≥60 years admitted to medical wards with HARRPE score of ≥0.2. They also assess elderly patients by proactive screening. In addition, patients can be referred to link nurses using a standardised clinical referral form. The form can be completed by any member of the clinical team, including doctors, nurses, pharmacists, and any allied health care professional. The criteria for clinical referral include items such as frequent readmission, poor social support, inadequate care at home, deterioration in memory, drug compliance problems, repeated falls, mobility, and functional impairment. Referrers can also comment about any problem not listed in the referral form. In HKWC, case recruitment and discharge planning take place in the medical wards of the acute hospital, Queen Mary Hospital (QMH), and three convalescent hospitals, namely the Fung Yiu King Hospital (FYKH), Grantham Hospital (GH), and Tung Wah Hospital (TWH). After assessment, link nurses will, according to need, allocate patients to either ICM Case Management or Home Support Team (HST) services (see below). In order to enhance the care of high-risk older patients in QMH, a Comprehensive Care Program for the Elderly (CCPE) area has been established. There are 12 beds designated as CCPE (6 male and 6 female beds) in paired wards of QMH. Case recruitment for CCPE is mainly from the AED. Patients in CCPE are under the care of the regular medical team with proactive ICDS multidisciplinary input including geriatric assessment and discharge planning for appropriate community support services.

In the community component, there are two important streams, namely the ICM Case Management and HST service. In ICM Case Management, each high-risk older patient is followed up by a case manager for a period of around 3 months following hospital discharge. In HKWC, in terms of full-time equivalence, two social workers, one physiotherapist (PT), one occupational therapist (OT), and half a nurse (advanced practice nurse) take turns to be a case manager. Case managers provide post-discharge support to older patients by home visits and telephone support. They are responsible for community service coordination and ensuring patient compliance with planned services and management. The second stream is the HST service and is the responsibility of a non-governmental organisation (NGO) partner. In HKWC, the NGO partner is Aberdeen Kai-fong Association (香港仔坊會). This HST includes nurses, PT, OT, and other allied health members. They provide rapid and intensive community support for discharged patients, offering services such as meal delivery, household cleaning, respite care, and home assessment and modification.

Case selection and allocation to ICM Case Management or HST is performed by link nurses under the supervision of an ICM geriatrician. Link nurses apply standardised selection criteria for case allocation. In general, patients with more complex medical and social problems who require multidisciplinary intervention by nurses, PT, and/or OT will be allocated to ICM Case Management. Those who require urgent social services are recruited into HST. Link nurses, ICM case managers, and the HST hold weekly multidisciplinary case conferences chaired by an ICM geriatrician. If needed, referral for rehabilitation in a geriatric day hospital, fast track clinic, or early specialist clinic follow-up can be offered to patients.

The ICDS programme that started in HKWC in 2012 is unprecedented and deserves a large-scale study to demonstrate its efficacy. Although the aim of ICDS is not to reduce costs, its value and sustainability can nonetheless be better justified if this can be achieved.

The objectives of this prospective cohort study were to investigate whether the ICDS can reduce AED attendance, acute hospital admissions, and hospital bed days (acute and convalescence), and to identify the independent factors that predict its efficacy. In addition, we wished to determine whether there is potential for ICDS to reduce health care costs.

This was a prospective cohort study performed in four hospitals of HKWC, namely the QMH, FYKH, GH, and TWH. The study protocol was approved by the Institutional Review Boards of the University of Hong Kong and HA HKWC.

Our subjects were home-dwelling older patients aged ≥60 years admitted to the general medical wards of QMH and were recruited into the ICDS programme by link nurses from 1 April 2012 to 30 March 2013. Patients were excluded from the analysis if they died, entered residential care homes for the elderly (RCHEs), moved out of the cluster, or refused ICDS services before their first home visit.

Baseline data included demography, HARRPE score (if available), mode of feeding, continence status, presence of pressure sores, and use of an indwelling urinary catheter, nasogastric tube, or long-term oxygen. The chief problems at index admission of ICDS recruitment were noted. In addition, data on co-morbidities, number of medications, and baseline blood tests including haemoglobin, albumin, and creatinine levels were retrieved from the HA Clinical Management System (CMS). The Charlson Comorbidity Index (CCI) was used to quantify the burden of co-morbid diseases,15 and quantified according to International Classification of Diseases (ICD) coding in CMS.

Cognitive status was assessed on entry to and discharge from the ICDS programme using the Abbreviated Mental Test (AMT).16 The patients’ Modified Functional Ambulation Category scale (MFAC) and Barthel Index 20 (BI-20) status were also recorded.17 The mortality rate and institutionalisation rate of patients within 6 months of intake were calculated.

The Statistical Package for the Social Sciences (Windows version 18.0; SPSS Inc, Chicago [IL], US) was used in statistical analysis. Continuous variables were expressed as mean ± standard deviation. Independent t test was used to compare continuous variables of two different groups. Paired t test was used to compare the continuous variables within groups. Mann-Whitney test and Wilcoxon signed rank test were used when the continuous variables could not be assumed to be in normal distribution. Chi squared test and Fisher’s exact test were employed to compare categorical variables. The association between different variables with any AED attendance and no reduction in bed days was calculated using univariate logistic regression. The variables were gender, programmes entered, use of home oxygen, use of tube feeding, use of a Foley catheter, presence of wound, BI-20, AMT, MFAC, CCI, haemoglobin level, albumin level, creatinine level, and number of medications. Significant factors detected during univariate analysis were put into multivariate stepwise backward logistic regression. Statistical significance was inferred by a two-tailed P value of 0.05.

From 1 April 2012 to 30 March 2013, among 7268 hospital discharges, 1184 (16.3%) home-dwelling older patients aged ≥60 years were recruited to the ICDS in HKWC. Of these patients, 23 died, 32 were institutionalised, and 39 refused to join the ICDS before the first home visit. A total of 1090 patients entered into the study. The baseline characteristics, demography, CCI, and chief problems at index admission of ICDS recruitment are shown in Table 1. Details of the programme entered are shown in Table 2. Table 3 illustrates the change in AED attendances, acute hospital admissions, and bed days (acute and convalescent hospitals) after joining the ICDS. Within 6 months of ICDS service, 85 (7.8%) patients died and only 26 (2.4%) older patients required institutionalisation in RCHEs. We observed a 40% reduction in AED attendances 6 months after initiation of the ICDS compared with 6 months before (mean, 1.2 vs 2.0 episodes; P<0.001). There was also a 47% reduction in acute hospital admissions (mean, 0.9 vs 1.7 episodes; P<0.001), and a 31% reduction in bed days (acute and convalescence) [mean, 11.1 vs 16.1 days; P<0.001] 6 months after joining the ICDS (Table 3).

There was mild improvement in MFAC and BI-20 on discharge from the ICDS compared with the level at entry (MFAC, 6.3 ± 2.2 vs 5.7 ± 1.6, P<0.001; BI-20, 17.6 ± 4.1 vs 16.5 ± 4.1, P<0.001). There was no significant change in AMT (8.4 ± 1.7 vs 8.4 ± 2.1; P=0.831).

Among 1090 subjects included, 596 (54.7%) required AED attendance within 6 months of joining the ICDS. Increasing age (odds ratio [OR]=1.019; confidence interval [CI], 1.002-1.036; P=0.025) and high CCI score (OR=1.178; 95% CI, 1.108-1.254; P=0.001) were independent positive predictors for AED attendance. A high albumin level (OR=0.957; 95% CI, 0.935-0.980; P=0.001) and living alone (OR=0.677; 95% CI, 0.473-0.969; P=0.033) were negative predictors for AED attendance (Table 4). Overall, 310 (28.4%) subjects had no reduction in bed days when comparing 6 months before and 6 months after joining the ICDS. Increasing age (OR=1.019; 95% CI, 1.002-1.036; P=0.025) and increasing number of medications (OR=1.062; 95% CI, 1.020-1.105; P=0.003) were significant independent positive predictors of no reduction in bed days; while higher haemoglobin level (OR=0.931; 95% CI, 0.871-0.995; P=0.034) was a negative predictor (Table 4).

In HKWC, the cost per patient day was HK$4461 for an acute-hospital medical bed and HK$2237 in a convalescent hospital. Each AED attendance also incurred a cost of HK$877. The average LOS in an acute medical ward at QMH was 2.4 days. The total number of bed days saved for acute hospitals were (1.7–0.9) x 1090 x 2.4 = 2093 days. In terms of acute-hospital bed days, the total cost-saving 6 months following ICDS compared with 6 months before was 2093 x $4461 = HK$9 336 873 (around HK$9.3 million). The total cost-saving for convalescent-hospital bed days in 6 months was ([16–11] x 1090 – 2093) x $2237 = HK$7 509 609 (around HK$7.5 million). The cost-saving due to reduced AED attendance was (2.0–1.2) x 1090 x $877 = HK$764 744 (around HK$0.76 million). The annual expenditure for the ICDS was HK$12.62 million in total, giving a net cost-saving over 6 months of (9.3+7.5+0.76) – 12.62/2 = HK$11.25 million. The potential annual cost-saving was thus HK$11.25 million x 2 = HK$22.5 million (approximately US$2.9 million).

This study demonstrates that the present ICDS in HKWC reduces AED attendance, hospital admissions, and hospital bed days. A local study has shown that medical patients are in general prone to institutionalisation following discharge from hospital.18 This programme appeared to keep older patients at home as evidenced by the low institutionalisation rate (2.4%). Nonetheless there was selection bias as those who required RCHE admission direct from hospital were excluded from the study. Hence, ICDS provided intervention in a group of older patients who had no imminent need of institutionalisation at the time of hospital discharge.

Different strategies have been described to reduce readmissions, namely geriatric assessment, discharge planning, a case manager approach, post-discharge support services, early intervention for ad-hoc medical problems, and use of telephone nursing services.6 7 8 9 10 11 One important element of success is a targeted approach that provides services to high-risk patients.12 In 2010, the Cochrane Database of Systematic Reviews revealed that a structured discharge plan tailored to the individual patient was likely to reduce hospital LOS and readmission rates for older people.19 The efficacy of a post-discharge programme that comprises the above elements was supported by another meta-analysis.20 The ICDS in HKWC comprises all the above elements with a targeted approach that focuses on high-risk older patients as identified by their HARRPE score. In addition, high-risk older patients who were not identified by a HAPPRE score were recruited as a result of link nurse screening and clinical referrals in hospital.

Older patients recruited in the ICDS with motor and functional problems underwent PT and OT assessment during home visits. Home exercise could be taught as appropriate with selected patients referred to the geriatric day hospital for rehabilitation. This may help explain why the ICDS was able to improve the functional and ambulatory status of patients. We observed no significant AMT change in our recruited patients upon closure of ICDS. The ICDS aims at maintaining older patients in the community during the high-risk period rather than improving their cognitive function.

Multivariate analysis revealed that increasing age, low albumin level, and high CCI score were associated with AED attendance 6 months after joining the ICDS. This result was very similar to that of a systematic review of the general risk factors for preventable readmissions.21 It concluded that increasing age and poor health as measured by CCI were associated with high readmission risk.21 Low serum albumin level is known to associate with poorer clinical and rehabilitation outcomes in older patients.22 In this study, patients with advanced age, low albumin level, and high CCI score were more likely to attend AED again, even after joining the ICDS programme. Based on these results, we may consider adjusting our programme to target these ‘ultra high-risk’ groups. In this study, living alone was a protective factor for AED attendance 6 months after joining the ICDS. Although previous studies showed that living alone was a risk factor for hospital admission, we found that this group of patients had significantly fewer AED attendances after ICDS.23 24 Indeed, age, albumin level and CCI were all better in the living alone group compared with those who were not. Living alone remained a protective factor after multivariate analysis with the above-mentioned factors adjusted, indicating that it was an independent predictor by itself. There are several possible explanations for this observation. First, older people living at home alone belong to a selected group who are usually more self-reliant. With home visits and telephone support by ICDS case managers or HST, together with geriatrician backup, they have a dependable team from whom advice can be sought for ad-hoc problems. On the contrary, those living with their family might be less independent. Their health-seeking behaviour is strongly influenced by family members or carers at home, who may prompt them or bring them to the AED for urgent consultations.

In this study, there was a group of patients for whom there was no reduction in bed days 6 months after joining the ICDS. These patients were in more advanced age, with a low haemoglobin level, and prescribed an increasing number of drugs. It is possible that the outcome of the ICDS may be further improved by correcting the anaemia and reducing polypharmacy in this group of older patients.

There were several limitations in this study. The study was not a randomised controlled trial. The ICDS programme is a government-funded service provided to all suitable older patients in Hong Kong. Hence, it was practically impossible to have a control group in the study. Bias in determining patients’ discharge time might occur as there was no blinding of treating doctors in the programme. The decision and time to discharge was subjective and could have been affected by health care workers who wanted the programme to demonstrate beneficial results. The link nurses in the programme could not perform discharge planning in 100% of the admitted high-risk patients, as their service was limited on public holidays and Sundays. In addition, only 16.3% of patients were recruited to the programme after discharge planning. These factors might have led to selection bias in the study. Seasonal variation in hospital admissions among high-risk older patients might also affect the validity of the results. Nonetheless, our patients joined the ICDS at different time points during recruitment and this, to a certain extent, may minimise the seasonal variation effect on hospital admission analysis. Statistically, no adjustment was made for AED attendance and hospital admissions for patients who were institutionalised or who died during the post–6-month period. The reduced number of patients due to deaths was also not considered during the end of study analysis. The appropriate analysis would use Cox’s regression, making use of the time to event (AED attendance), and subjects who died or were institutionalised during the follow-up period would be censored and not just excluded from the analysis. Since the index admission was counted as the pre-ICDS period, unplanned admission during the pre-ICDS period would start from ‘one’, putting the performance during that period at a great disadvantage when compared with the post-ICDS period. This study only looked at patients admitted to hospitals in HKWC. This limited the generalisability of the results of ICDS in other clusters. The CCI used in this study was quantified based on ICD coding in CMS and might have led to undercoding and consequent underestimation of CCI for patients. Although there was a potential annual cost-saving of HK$22.5 million with the ICDS, this was just a crude calculation and saving based on reduced bed days: reduced AED attendance provided a nominal saving. There was no actual reduction in staff requirements and other expenses as a result of reduced admissions. In addition, we did not consider planned readmissions that also contributed to health care expenditure. The ICDS is designed to prevent unplanned readmission. The ICM case managers and HST rarely interfere with planned readmission for patients, apart from reminding them to follow the schedule. Thus analysis of the planned readmissions might not have impacted greatly on our study findings.

The ICDS reduces AED attendance, unplanned acute hospital admissions, and hospital bed days in high-risk older patients. Additional studies are suggested to determine whether further reductions can be achieved by modifying some of the predictive factors identified in this study. A more detailed auditing is also warranted to demonstrate the value of ICDS in reducing health care costs.

1. Gordon A. ‘Revolving door syndrome’. Elder Care 1995;7:9-10,12.

2. Jencks SF, Williams MV, Coleman EA. Rehospitalizations among patients in the Medicare fee-for-service program. N Engl J Med 2009;360:1418-28. Crossref

3. Balla U, Malnick S, Schattner A. Early readmissions to the department of medicine as a screening tool for monitoring quality of care problems. Medicine (Baltimore) 2008;87:294-300. Crossref

4. Wong EL, Cheung AW, Leung MC, et al. Unplanned readmission rates, length of hospital stay, mortality, and medical costs of ten common medical conditions: a retrospective analysis of Hong Kong hospital data. BMC Health Serv Res 2011;11:149. Crossref

5. Luk JK, Tuet OS, Chan FH. Unplanned readmission among older medical patients discharged from an extended care hospital. International Hospital Federation Convention; 2001 May 15-18; Hong Kong.

6. Luk JK, Or KH, Woo J. Using the comprehensive geriatric assessment technique to assess elderly patients. Hong Kong Med J 2000;6:93-8.

7. Naylor MD, Brooten D, Campbell R, et al. Comprehensive discharge planning and home follow-up of hospitalized elders: a randomized clinical trial. JAMA 1999;281:613-20. Crossref

8. Hyde CJ, Robert IE, Sinclair AJ. The effects of supporting discharge from hospital to home in older people. Age Ageing 2000;29:271-9. Crossref

9. Phillips CO, Wright SM, Kern DE, Singa RM, Shepperd S, Rubin HR. Comprehensive discharge planning with postdischarge support for older patients with congestive heart failure: a meta-analysis. JAMA 2004;291:1358-67. Crossref

10. Steward S, Pearson S, Horowitz JD. Effects of a home-based intervention among patients with congestive heart failure discharged from acute hospital care. Arch Intern Med 1998;158:1067-72. Crossref

11. Lim WK, Lambert SF, Gray LC. Effectiveness of case management and post-acute services in older people after hospital discharge. Med J Aust 2003;178:262-6.

12. Coleman EA, Parry C, Chalmers S, Min SJ. The care transitions intervention: results of a randomized controlled trial. Arch Intern Med 2006;166:1822-8. Crossref

13. Chan S, Kwong P, Kong B, et al. Improving Health of High Risk Elderly in the Community—the HARRPE. Proceedings of the Symposium on Community Engagement III “Creating a Synergy for Community Health” 2008; Hong Kong. Hong Kong: Hong Kong Hospital Authority; 2008.

14. Ng MF, Sha KY, Tong BC. Bridging the gap: win-win from Integrated Discharge Support for Elderly Patients. HA Convention; 2011.

15. Chan TC, Luk JK, Chu LW, Chan FH. Validation study of Charlson Comorbidity Index in predicting mortality in Chinese older adults. Geriatr Gerontol Int 2013;14:452-7. Crossref

16. Chu LW, Pei CK, Ho MH, Chan PT. Validation of the Abbreviated Mental Test (Hong Kong version) in the elderly medical patients. Hong Kong Med J 1995;1:207-11.

17. Mahoney FI, Barthel DW. Functional evaluation: The Barthel index. Md State Med J 1965;14:61-5.

18. Luk JK, Chiu PK, Chu LW. Factors affecting institutionalization in older Chinese patients after recovery from acute medical illnesses. Arch Geront Geriatr 2009;49:e110-4. Crossref

19. Shepperd S, McClaran J, Phillips CO, et al. Discharge planning from hospital to home. Cochrane Database Syst Rev 2010;(1):CD000313.

20. Fox MT, Persaud M, Maimets I, Brooks D, O’Brien K, Tregunno D. Effectiveness of early discharge planning in acutely ill or injured hospitalized older adults: a systematic review and meta-analysis. BMC Geriatr 2013;13:70. Crossref

21. Vest JR, Gamm LD, Oxford BA, Gonzalez MI, Slawson KM. Determinants of preventable readmissions in the United States: a systematic review. Implement Sci 2010;5:88. Crossref

22. Luk JK, Chiu PK, Tam S, Chu LW. Relationship between admission albumin levels and rehabilitation outcomes in older patients. Arch Gerontol Geriatr 2011;53:84-9. Crossref

23. Arbaje AI, Wolff JL, Yu Q, Powe NR, Anderson GF, Boult C. Postdischarge environmental and socioeconomic factors and the likelihood of early hospital readmission among community-dwelling Medicare beneficiaries. Gerontologist 2008;48:495-504. Crossref

24. Hanania NA, David-Wang A, Kesten S, Chapman KR. Factors associated with emergency department dependence of patients with asthma. Chest 1997;111:290-5. Crossref

The use of endoscopic resection in the treatment of superficial gastro-intestinal neoplasms is gaining popularity worldwide.1 2 3 4 5 6 7 8 9 10 11 Endoscopic mucosal resection (EMR) and endoscopic submucosal dissection (ESD) are the key endoscopic methods pioneered in Japan and Korea for resection of these lesions.1 2 3 4 5 6 7 Early gastric or oesophageal neoplasms are associated with a low rate of lymph node metastasis and endoscopic resection of these neoplasms has been shown to be associated with a low rate of morbidity and high long-term survival.6 7 Nevertheless, the adoption of these techniques outside high-volume countries has been slow, as the recognition of early gastric and oesophageal neoplasms is difficult and the procedures required for endoscopic resection are technically demanding and have a long learning curve.8 9 10 11 12 13 14

In Hong Kong, an increased awareness and recognition of early gastro-intestinal neoplasms have resulted in greater frequency of diagnosis of these lesions. Hence advanced endoscopic resection procedures such as EMR or ESD are also being performed increasingly. However, the scope of these services provided by government-funded hospitals operated by the Hospital Authority of Hong Kong has not been previously defined. In addition there are limited data on the outcome of these advanced endoscopic procedures outside high-volume centres of Japan and Korea.

The aim of the current study was to perform a region-wide audit of the short-term clinical and oncological outcomes of EMR and ESD for superficial upper gastro-intestinal neoplasms in all government-funded hospitals in Hong Kong.

This was a Hong Kong–wide retrospective review commissioned by the Hospital Authority of Hong Kong of all patients who underwent EMR or ESD for superficial upper gastro-intestinal lesions between January 2010 and June 2013 in 12 government-funded hospitals. Patients were identified using the Clinical Data Analysis and Reporting System (CDARS) based on their diagnostic and procedural coding (9th edition of the International Classification of Diseases). The CDARS is a computer-based administration database that records all the diagnostic and procedural coding of admitted patients. Patient data were retrieved and reviewed manually through the system. Data were recorded for indication of the procedures, peri-procedural and procedural parameters, oncological outcomes, and morbidity and mortality. The study was performed according to the Declaration of Helsinki.

All patients who underwent EMR or ESD of the oesophagus, stomach, and duodenum were included in the study. To differentiate these patients from those that received simple polypectomy, EMR was defined as the act of performing mucosectomy with prior injection of a cushioning fluid into the submucosa to hasten removal of the lesions. It was acknowledged that EMR could be performed by a variety of methods but this review did not attempt to sub-classify patients according to the different techniques used.1 On the other hand, ESD was defined as a more refined form of mucosectomy that involved submucosal injection of a cushioning fluid, circumferential incision of the mucosa, followed by dissection of the submucosal tissue to free the lesion away from surrounding tissue. The type of endoscopic knives used for mucosal incision and submucosal dissection were also recorded.

The outcomes of EMR and ESD were assessed clinically and oncologically.

Compared with conventional polypectomy, ESD is a technically demanding procedure that is associated with an increased risk of complications. In particular, the risk of perforation and bleeding are heightened in inexperienced operators.13 14 In the current study, intra-procedural and post-procedural complications were recorded and factors alluding to development of these complications were also reviewed.

To determine whether EMR or ESD can provide adequate oncological clearance for neoplastic lesions, assessment of long-term survival is essential. However, since EMR or ESD procedures have been performed in Hong Kong for only a short period of time, such assessment was not possible. Thus, the adequacy of oncological control of EMR or ESD was based on completeness of resection as judged during the procedure, presence of margin involvement (lateral and deep), the need for piecemeal resection, and recurrence rates. En-bloc resection was defined as resection of the tumour in one piece. R0 resection was defined as resection of the tumour with clear lateral and vertical margins. R1 resection was defined as microscopic involvement of the margins. R2 resection was defined as macroscopic involvement of margins as noted during endoscopy. Local recurrence was defined as recurrent cancer detected at the primary resection site during follow-up oesophagogastroduodenoscopy when pathological review of the ESD specimen revealed no tumour on the lateral and vertical margins.

Statistical analyses were done mainly using descriptive analysis. The predictors of morbidity and local recurrences were analysed by multivariate logistic regression analysis using the following factors: sex, American Society of Anesthesiologists grading, presence of neoplastic lesions (adenoma, dysplasia, carcinoma), ESD performed as a staging procedure, the need for piecemeal resection, pathological size, lateral and vertical margin involvement, depth of pathological invasion, R0 resection, and the institution volume of ESD. The 2-year overall and disease-specific survivals were calculated using the Kaplan-Meier estimator. A two-sided P value of <0.05 was considered statistically significant. Statistical analyses of data were performed using the Statistical Package for the Social Sciences (Windows version 20.0; SPSS Inc, Chicago [IL], US).

During the study period, 187 lesions in 168 patients were resected. The patient demographics are shown in Table 1. The mean (± standard deviation) age of patients was 61.9 ± 14.3 years and 61.3% were male. Overall, EMR was performed in 34 (18.2%) lesions and ESD in 153 (81.8%) lesions for dysplasia or carcinoma with the majority of lesions located in the stomach (132 lesions, 70.6%), followed by the oesophagus (43 lesions, 23%) and the duodenum (12 lesions, 6.4%). The distribution in the numbers of the procedures among various hospitals is shown in Figure a.

The clinical outcomes of the procedures are shown in Table 2. General anaesthesia with tracheal intubation was required for the majority of procedures (n=148, 79.1%). The procedural times were recorded in 50 (26.7%) procedures only and were not representative of the actual time required for the procedures. The mean size of the lesions was 2.6 ± 1.8 cm. The 30-day morbidity rate was 14.4%. Intra-procedural complications occurred in 13 (7.0%) procedures, with one procedure having both bleeding and perforation. Perforations occurred in eight (4.3%) procedures and all were controlled with endoscopic clipping. Severe bleeding requiring endoscopic clipping or transfusion occurred in six (3.2%) procedures. One procedure required conversion to open surgery due to uncontrolled bleeding. Post-procedural complications occurred in 14 (7.5%) procedures; the most common causes of which were bleeding (4.3%), stricture formation (1.6%), and perforations (1.1%). All post-procedural bleeding and stricture formation were managed endoscopically. The procedures having perforation were managed conservatively. No patients had post-procedural mortality. The mean hospital stay was 3.7 ± 3.1 days. Haemoglobin level was checked on the first post-procedural day in 116 (62%) procedures. Proton pump inhibitors were prescribed to 171 (91.4%) procedures; of these, 93 (80.2%) were administered to ESD in the stomach. No difference in size of lesions, hospital stay, or morbidities was observed between university– and non-university–affiliated hospitals.

The types of knives used for mucosal incision included the Dual knife (35.8%; KD-650U, Olympus Co Ltd, Tokyo, Japan), the triangular tip knife (14.4%; KD-640L, Olympus Co Ltd, Tokyo, Japan), the needle knife (8%; KD-1L-1, Olympus Co Ltd, Tokyo, Japan), the insulated tip (IT2) knife (7%; KD-610L, Olympus Co Ltd, Tokyo, Japan), and the Hybrid knife (3.7%; ERBE Tübingen, Germany). Knives used for submucosal dissection included the Dual knife (48%; KD-650U, Olympus Co Ltd), the insulated tip (IT2) knife (24.6%; KD-610L, Olympus Co Ltd), the triangular tip knife (11.2.%; KD-640L, Olympus Co Ltd), and the Hybrid knife (4.3%; ERBE).

The oncological outcome of the procedures is shown in Table 3. In 22 (11.8%) lesions, en-bloc resection was not possible and the lesions had to be resected in piecemeal. Among patients having adenoma, dysplasia, or carcinoma (n=100), R0 resection was achieved in 78%. In 7% of the patients, pre-procedural investigations were inconclusive on the feasibility of endoscopic resection and ESD was performed as a staging procedure. Lateral margin involvement occurred in 14 (14%) procedures and vertical margin involvement occurred in eight (8%) procedures. Local recurrences occurred in nine (9%) patients and five were treated by further ESD. Residual disease was present in 15 (15%) patients, four due to failure to complete ESD and 11 due to margin involvement. Of these 15 patients with residual disease, complete resection was achieved in eight during salvage surgical resection, two patients underwent repeat endoscopic resection, two underwent radiofrequency ablation, and three refused treatment. The mean follow-up time of patients was 20.6 ± 10.8 months and the 2-year overall survival rate and disease-specific survival rate was 90.6% and 100%, respectively (Fig b). Overall, 38.5% of patients were followed up for at least 2 years. No differences in oncological outcomes were observed between university– and non-university–affiliated hospitals.

The predictors of morbidity in patients who received endoscopic resection were then analysed with multivariate logistic regression (Table 4). Specimen size of ≥5 cm (P=0.001), the need for piecemeal resection (P=0.032), and pathological invasion to the submucosa or deeper (P=0.042) were independent predictors of morbidity. The predictors for local recurrences were also analysed: the need for piecemeal resection was the only independent predictor (P=0.011; Table 5).

Using a hospital admission database, the outcome for patients who underwent endoscopic resection of superficial upper gastro-intestinal neoplasms among 12 government-funded hospitals in Hong Kong was reviewed. The total number of procedures performed was higher than expected and the procedures were associated with a low risk of morbidity. Nonetheless there may be room for improvement in the oncological outcome, particularly in terms of rates of margin involvement and the number of patients with residual disease. The predictors of morbidity and local recurrence were in line with those reported from high-volume centres and the 2-year survival of patients who had adenoma, dysplasia, or carcinoma was excellent following endoscopic resection.

The adoption of ESD outside Japan and Korea has been slow.5 8 15 16 17 18 The reported studies were mostly small series and the results were variable. In four European and two South-East Asian studies, the number of patients included was between 23 and 70, the en-bloc resection rate ranged from 25% to 100% and morbidity 0% to 24%. These results contrast with those obtained from Japan and Korea. In the stomach, the rate of en-bloc resection was 94.9% to 95.3% and piecemeal resection was 4.1%. The risk of bleeding was 1.8% to 16.2% and perforation was 1.2% to 4.5%. The risk of recurrence was less than 1% and the 3- and 5-year overall survival rates were 98.4% and 97.1% respectively.6 7 These results reflect the difficulty of introducing a technically demanding ESD programme in low-to-medium volume localities outside Japan and Korea.19 The reasons for poorer clinical and oncological outcomes may be explained by a combination of factors. These include the early experience and learning curve issues, technical difficulties leading to the need for piecemeal resection, and failure to recognise tumour margins.

The current study is the largest non-Japanese and non-Korean report on the outcomes of ESD. The results illustrate that an acceptable en-bloc resection rate and morbidity rate can be achieved in low-to-medium–volume hospitals. The introduction of endoscopic resection techniques in Hong Kong is in its infancy with many challenges similar to those described in western literature.20 21 22 The emphasis in our region, however, was not just on the technical performance of ESD but also the ability to detect and diagnose these lesions. In 2011, a total of 1101 new cases of gastric cancer were diagnosed in Hong Kong but the percentage of cases that were amenable to endoscopic resection is unknown.23 Based on the findings of this study, the percentages are likely to have been much less than 10%. This figure contrasts significantly with those reported from Spain (20%) and Japan (53%).24 25 Hence, measures to further improve the early diagnosis of upper gastro-intestinal malignancies are needed. Key personnel to perform ESD in individual hospitals should be identified and receive intensive training in screening endoscopy to improve detection of early lesions.26 These individuals should attend local workshops and clinical attachments at high-volume centres in Japan and Korea to acquire such skills. Enhanced endoscopic imaging systems (narrow band imaging, flexible spectral imaging colour enhancement, or autofluorescence imaging) that may improve the ease of diagnosing early upper gastro-intestinal lesions should also be obtained by those hospitals interested in developing the technique and these should be supported by the government on a regional scale.27 28 29 The use of population-based screening programmes in our locality may not be justified because of the moderate incidence of upper gastro-intestinal cancers, but studies to evaluate screening of high-risk groups may be justified.30 31 Better recognition of early upper gastro-intestinal neoplasms not only increases the detection rates but may also allow more accurate pre-ESD assessment and improve the rates of margin involvement and oncological outcomes.

It is acknowledged that ESD remains a technically demanding procedure that is associated with risk of perforation and bleeding. To master such skills, the surgeon must be familiar not only with the ESD procedure, but also with the methods used to treat complications. In addition, the difficulty of ESD depends on the organ involved, as well as the location within that organ.32 33 Thus, in high-volume centres, learning of ESD is usually done in a stepwise approach, starting from easy small lesions located at the antrum of the stomach and progressing to more difficult lesions or lesions located in other parts of the gastro-intestinal tract.34 35 The number required for mastering the technique is 20 to 40 procedures in each location. These numbers are unlikely to be attainable within a short period of time in Hong Kong and most western countries. Hence, each locality will need to decide on the most appropriate training strategy to overcome the learning curve: this will likely involve intensive training with the use of animal models, apprenticeship with local experts, and overseas training in high-volume centres.13 20 24 It may also be beneficial to the region to restrict the performance of ESD to a few key personnel in the main institutions and once they are proficient, involve other endoscopists in the process.

To further improve the outcome of endoscopic resection of upper gastro-intestinal lesions in Hong Kong, a number of measures should be implemented. The establishment of a task force on endoscopic diagnosis and treatment of early gastro-intestinal cancers with regular training should be considered. A local standard in endoscopic reporting and pathological examination is required as these assessments of early upper gastro-intestinal neoplasms are unique and pivotal to guiding treatment. A ‘best practice’ guideline for performing ESD and the provision of pre- and post-ESD care should be established to provide a benchmark for future audits.

There were a number of limitations to the current study. First, patients included in the study were identified using CDARS based on the diagnosis and procedural coding. There is a risk of unidentified procedures if they were incorrectly coded. Also, those procedures that were performed outside the Hospital Authority hospitals could not be accounted for. Second, since this is a retrospective review, some outcome parameters might not have been properly defined or available for all patients. Finally, since there is no standardised reporting system in Hong Kong for ESD or EMR and pathological assessment, there was a risk of information bias during extraction of the data.

The introduction of ESD in Hong Kong is still in its infancy; EMR and ESD were introduced in low-to-moderate–volume hospitals with acceptable morbidity rates. The short-term survival was excellent. Nonetheless, other oncological outcomes were higher than those observed in high-volume centres and more secondary procedures were required.

1. Soetikno R, Kaltenbach T, Yeh R, Gotoda T. Endoscopic mucosal resection for early cancers of the upper gastrointestinal tract. J Clin Oncol 2005;23:4490-8. Crossref

2. Oka S, Tanaka S, Kaneko I, et al. Advantage of endoscopic submucosal dissection compared with EMR for early gastric cancer. Gastrointest Endosc 2006;64:877-83. Crossref

3. Fujishiro M, Yahagi N, Kakushima N, et al. Endoscopic submucosal dissection of esophageal squamous cell neoplasms. Clin Gastroenterol Hepatol 2006;4:688-94. Crossref

4. Fujishiro M, Yahagi N, Kakushima N, et al. Outcomes of endoscopic submucosal dissection for colorectal epithelial neoplasms in 200 consecutive cases. Clin Gastroenterol Hepatol 2007;5:678-83. Crossref

5. Chiu PW, Chan KF, Lee YT, Sung JJ, Lau JY, Ng EK. Endoscopic submucosal dissection used for treating early neoplasia of the foregut using a combination of knives. Surg Endosc 2008;22:777-83. Crossref

6. Isomoto H, Shikuwa S, Yamaguchi N, et al. Endoscopic submucosal dissection for early gastric cancer: a large-scale feasibility study. Gut 2009;58:331-6. Crossref

7. Chung IK, Lee JH, Lee SH, et al. Therapeutic outcomes in 1000 cases of endoscopic submucosal dissection for early gastric neoplasms: Korean ESD Study Group multicenter study. Gastrointest Endosc 2009;69:1228-35. Crossref

8. Probst A, Golger D, Arnholdt H, Messmann H. Endoscopic submucosal dissection of early cancers, flat adenomas, and submucosal tumors in the gastrointestinal tract. Clin Gastroenterol Hepatol 2009;7:149-55. Crossref

9. Dinis-Ribeiro M, Pimentel-Nunes P, Afonso M, Costa N, Lopes C, Moreira-Dias L. A European case series of endoscopic submucosal dissection for gastric superficial lesions. Gastrointest Endosc 2009;69:350-5. Crossref

10. Repici A, Hassan C, Carlino A, et al. Endoscopic submucosal dissection in patients with early esophageal squamous cell carcinoma: results from a prospective Western series. Gastrointest Endosc 2010;71:715-21. Crossref

11. Probst A, Golger D, Anthuber M, Märkl B, Messmann H. Endoscopic submucosal dissection in large sessile lesions of the rectosigmoid: learning curve in a European center. Endoscopy 2012;44:660-7. Crossref

12. Kakushima N, Fujishiro M, Kodashima S, Muraki Y, Tateishi A, Omata M. A learning curve for endoscopic submucosal dissection of gastric epithelial neoplasms. Endoscopy 2006;38:991-5. Crossref

13. Teoh AY, Chiu PW, Wong SK, Sung JJ, Lau JY, Ng EK. Difficulties and outcomes in starting endoscopic submucosal dissection. Surg Endosc 2010;24:1049-54. Crossref

14. Berr F, Ponchon T, Neureiter D, et al. Experimental endoscopic submucosal dissection training in a porcine model: learning experience of skilled Western endoscopists. Dig Endosc 2011;23:281-9. Crossref

15. Rösch T, Sarbia M, Schumacher B, et al. Attempted endoscopic en bloc resection of mucosal and submucosal tumors using insulated-tip knives: a pilot series. Endoscopy 2004;36:788-801. Crossref

16. Farhat S, Chaussade S, Ponchon T, et al. Endoscopic submucosal dissection in a European setting. A multi-institutional report of a technique in development. Endoscopy 2011;43:664-70. Crossref

17. Schumacher B, Charton JP, Nordmann T, Vieth M, Enderle M, Neuhaus H. Endoscopic submucosal dissection of early gastric neoplasia with a water jet-assisted knife: a Western, single-center experience. Gastrointest Endosc 2012;75:1166-74. Crossref

18. Chang CC, Lee IL, Chen PJ, et al. Endoscopic submucosal dissection for gastric epithelial tumors: a multicenter study in Taiwan. J Formos Med Assoc 2009;108:38-44. Crossref

19. Chiu PW. Novel endoscopic therapeutics for early gastric cancer. Clin Gastroenterol Hepatol 2014;12:120-5. Crossref

20. Coman RM, Gotoda T, Draganov PV. Training in endoscopic submucosal dissection. World J Gastrointest Endosc 2013;5:369-78. Crossref

21. Neuhaus H. Endoscopic submucosal dissection in the upper gastrointestinal tract: present and future view of Europe. Dig Endosc 2009;21 Suppl 1:S4-6. Crossref

22. Deprez PH, Bergman JJ, Meisner S, et al. Current practice with endoscopic submucosal dissection in Europe: position statement from a panel of experts. Endoscopy 2010;42:853-8. Crossref

23. Hong Kong Cancer Registry, Hospital Authority. Available from: http://www3.ha.org.hk/cancereg/. Accessed 11 Feb 2013.

24. Miguélez Ferreiro S, Cornide Santos M, Martínez Moreno E. Gastric cancer in a Spanish hospital: Segovia General Hospital (2005-2008) [in Spanish]. Gastroenterol Hepatol 2012;35:684-90. Crossref

25. Sugano K. Gastric cancer: pathogenesis, screening, and treatment. Gastrointest Endosc Clin N Am 2008;18:513-22, ix. Crossref

26. Yamazato T, Oyama T, Yoshida T, et al. Two years’ intensive training in endoscopic diagnosis facilitates detection of early gastric cancer. Intern Med 2012;51:1461-5. Crossref

27. Muto M, Minashi K, Yano T, et al. Early detection of superficial squamous cell carcinoma in the head and neck region and esophagus by narrow band imaging: a multicenter randomized controlled trial. J Clin Oncol 2010;28:1566-72. Crossref

28. Jung SW, Lim KS, Lim JU, et al. Flexible spectral imaging color enhancement (FICE) is useful to discriminate among non-neoplastic lesion, adenoma, and cancer of stomach. Dig Dis Sci 2011;56:2879-86. Crossref

29. Lee JH, Cho JY, Choi MG, et al. Usefulness of autofluorescence imaging for estimating the extent of gastric neoplastic lesions: a prospective multicenter study. Gut Liver 2008;2:174-9. Crossref

30. Takenaka R, Kawahara Y, Okada H, et al. Narrow-band imaging provides reliable screening for esophageal malignancy in patients with head and neck cancers. Am J Gastroenterol 2009;104:2942-8. Crossref

31. Leung WK, Wu MS, Kakugawa Y, et al. Screening for gastric cancer in Asia: current evidence and practice. Lancet Oncol 2008;9:279-87. Crossref

32. Oda I, Odagaki T, Suzuki H, Nonaka S, Yoshinaga S. Learning curve for endoscopic submucosal dissection of early gastric cancer based on trainee experience. Dig Endosc 2012;24 Suppl 1:129-32. Crossref

33. Murata A, Okamoto K, Muramatsu K, Matsuda S. Endoscopic submucosal dissection for gastric cancer: the influence of hospital volume on complications and length of stay. Surg Endosc 2014;28:1298-306. Crossref

34. Choi IJ, Kim CG, Chang HJ, Kim SG, Kook MC, Bae JM. The learning curve for EMR with circumferential mucosal incision in treating intramucosal gastric neoplasm. Gastrointest Endosc 2005;62:860-5. Crossref

35. Gotoda T, Friedland S, Hamanaka H, Soetikno R. A learning curve for advanced endoscopic resection. Gastrointest Endosc 2005;62:866-7. Crossref

Injuries to the hand and fingers are extremely common in children, yet they can have a significant impact on a child’s growth and development. Fingers are used to explore surroundings and perform daily activities such as playing, eating, and homework. Restricting children from these activities due to injuries can have immediate short- and long-term detrimental effects on the function of the hand, psychological wellbeing, and quality of life of the children. A 10-year review on the psychological impact on children and adolescents with finger or hand injuries noted that “Hand injuries are common and loss of a dominant hand or opposition is most important [sic]. Self-esteem and skill are associated with hand sensation, appearance, and functions.”1

Studies by Al-Anazi2 and Doraiswamy3 have identified crushing injury of finger(s) by door as the main cause of finger injuries in children. However, there has been no local study to identify the main cause of finger injuries in Hong Kong. In 2007, Lau and Ho presented data on the epidemiology of childhood finger injuries (unpublished data; Lau M, Ho PC. 20th Annual Congress of the Hong Kong Society for Surgery of the Hand, Hong Kong, 2007) that supported the findings in other cities. Similar to Al-Anazi2 and Doraiswamy,3 Lau and Ho found that crushing injury of finger by door was the most common cause of paediatric finger injuries from 2003 to 2005, and recommended various preventive measures.

The present study aimed to compare the previous set of data from 2003 to 2005 reported by Lau and Ho with more recent data obtained from 2010 to 2012. By comparing the epidemiology and mechanisms of finger injuries among local Hong Kong children, we aimed to determine whether there have been any significant changes over the past 5 years.

Data of patients admitted to Prince of Wales Hospital from 1 January 2003 to 31 December 2005 (group A) and from 1 January 2010 to 31 December 2012 (group B) were retrieved using the Clinical Data Analysis and Reporting System (CDARS) of the Hospital Authority’s Clinical Management System. Children aged 0 to 16 years, and with at least one of the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) codes listed in the Box among the top three diagnoses were included in the analysis.

Discharge summaries of all patients were reviewed to identify the mechanisms of finger injuries. For children in whom the mechanism was not immediately discernable from the discharge summary, further clarifications were obtained by telephone interviews with the child’s parents, which were conducted in 2006 for group A and in 2013 for group B. For children in whom crushing injury of finger(s) were due to closing doors, additional data were collected by telephone interviews with their parents using a specifically designed questionnaire (Fig 1).

Group A consisted of 140 children who presented with finger injury to Prince of Wales Hospital from 1 January 2003 to 31 December 2005. Three children from this group were excluded due to coding error. Group B comprised 109 children who presented from 1 January 2010 to 31 December 2012. No children from this group were excluded.

In both groups, crushing injury of finger by door was the most common cause of injury—45 (33%) in group A and 45 (41%) in group B—followed by sports injury, cut, and slip and fall (Table 1). Among children with crushing injury of finger by door, younger children were the most commonly injured (Fig 2a). The male-to-female ratio was 1:1.25 in group A and 1.37:1 in group B.

In the telephone interviews conducted with the parents of the 45 children who had crushing injury of finger by door, parents of two children in group A and six children in group B could not be contacted. Overall, all the parameters measured were similar between the periods 2003-2005 and 2010-2012. In both groups, most of the fingers involved were from the right hand, with the middle, ring, and little fingers being more commonly affected than the other fingers (Fig 2b).

Most of the injuries occurred at home—29 (67%) in group A and 27 (69%) in group B. At home, fingers were most frequently crushed at the hinge side of the door—27 (63%) in group A and 25 (64%) in group B—followed by the lock side and the middle of a double door. The doors were frequently closed by another child—16 (37%) in group A and 9 (23%) in group B and, in more than half of the cases, occurred even in the presence of adults—26 (60%) in group A and 22 (56%) in group B (Table 2).

The types of injury and their relative frequencies were compared between the groups (Table 3). Among the more common injuries were: nailbed injury—22 (31%) in group A and 22 (39%) in group B; fracture—17 (24%) in group A and 13 (23%) in group B; and laceration—17 (24%) in group A and 11 (20%) in group B. Most of the children in both group A (31 [72%]) and group B (30 [77%]) required operation. Among the 31 operations in group A, 21 (68%) were performed under general anaesthesia. By contrast, only 12 (40%) of the 30 operations in group B involved general anaesthesia.

Clinical outcomes were assessed when the telephone interviews were conducted, ie, in 2006 for group A and in 2013 for group B. Most children had a good recovery following treatment. Overall, 42 (98%) of children in group A and 37 (94%) of children in group B reported no pain. Only minor cosmetic problems prevailed in most children, with 34 (79%) in group A and 35 (89%) in group B rating their current level of cosmesis over 7 out of 10 (score 10 = no cosmetic problem). The injuries had minimal adverse effects for most children, with 42 (98%) in group A and 39 (100%) in group B rated their daily activities with a score over 8 (score 10 = no problem) [Fig 3].

However, five (12%) children in group A and four (10%) children in group B had crushing injury of finger by door resulting in finger amputation. Altogether seven children in groups A and B received replantation or revascularisation, one child underwent open reduction and fixation only, and one had a failed replantation due to failure to locate the arteries intra-operatively.

In group A, one child developed thrombosis following replantation of the right ring finger, requiring a subsequent revascularisation procedure 3 days later. This was complicated by hooknail deformity 1 year post-replantation and was subsequently treated by further reconstructive procedures. The levels of satisfaction in terms of appearance and daily function at final follow-up were rated 5 and 3 (out of 10, with 10 means no cosmetic problem and no problem in daily activities), respectively.

Crushing injury of finger by door is common. The true incidence of this type of injury is likely to be higher, as our data were limited to public hospitals so relied on the correct entry of ICD-9-CM codes into the CDARS. Data from the accident and emergency department and private practitioners were not analysed. Furthermore, many minor injuries might have been managed at home and not reported.

Crushing injury of finger by door is not just a local problem. Studies from Saudi Arabia and Glasgow showed that this type of injury accounted for most childhood fingertip injuries in these areas.2 3 These injuries consistently occurred at home, with the involved finger being frequently crushed at the hinge side of doors. Younger children were mostly affected. The similarity in epidemiology between the overseas data and our local data can help with recommendations for suitable door safety devices.

In this study, we identified that crushing injury by door was the major cause of paediatric finger injuries leading to hospital admission in both 2003-2005 and 2010-2012. Although most children were satisfied with the level of pain, cosmesis, and daily function of the injured digit at their final follow-up after treatment, serious injuries involving fractures and amputations occurred in a minority of patients. In addition to the surgical intervention and long-term hospitalisation required, these injuries could further lead to detrimental effects on the children’s growth and development.

Our study showed the presence of adults did not reduce the rate of these accidents, since most occurred even in the presence of an adult. This highlights the need for other preventative measures.

Many types of safety devices are easily available and affordable in Hong Kong. As the hinge side of doors is the most common side for fingers to be crushed, finger guard devices can be installed to prevent fingers being trapped in the opposing surfaces. Triangular-shaped rubbers, plastic or wooden stoppers can be inserted at the bottom of a door to prevent spontaneous closure. Magnets applied to the back of a door and its opposing wall surface present another equally effective and simple method of preventing unintended door closures. Dampers can be set up to reduce the speed of closing doors, thereby decreasing the force exerted on trapped fingers. The use of automatic doors should be avoided.

Yet, despite the easy availability and accessibility of these safety devices, there has been no significant change or improvement in terms of incidence and morbidity of children with crushing injury of fingers by door admitted to Prince of Wales Hospital in the 5-year period between 2003-2005 and 2010-2012. Thus, we should promote public awareness about this type of injury and provide more educational programmes on safety precautions in order to reduce the incidence of crushing injury of finger by door.

Crushing injury of finger by door accounts for the most common cause of paediatric finger injury requiring hospitalisation in Hong Kong. These injuries frequently result in hospital admission and surgical intervention, with considerable morbidity and high treatment cost. Crushing injury of finger by door occurs even in the presence of adults. Despite the easily available and affordable preventative measures in Hong Kong, our comparison revealed no significant difference in the incidence, nature, and severity of these domestic injuries between the years 2003-2005 and 2010-2012. Thus, it is our view that more effort should be invested into raising public awareness and education about these preventable injuries and to promote prevention measures.

1. Stoddard F, Saxe G. Ten-year research review of physical injuries. J Am Acad Child Adolesc Psychiatry 2001;40:1128-45. Crossref

2. Al-Anazi AF. Fingertip injuries in paediatric patients—experiences at an emergency centre in Saudi Arabia. J Pak Med Assoc 2013;63:675-9.

3. Doraiswamy NV. Childhood finger injuries and safeguards. Inj Prev 1999;5:298-300. Crossref