Lower urinary tract symptoms (LUTS) collectively represent a common condition among ageing men.1 2 There are limited data on the frequency of individual symptoms in patients who seek specialist care. It is well established that LUTS have far reaching impacts on patients’ daily lives.3 4 There are standard tools available to assess the severity of LUTS, such as the International Prostate Symptom Score, but they do not reflect the degree to which patients are subjectively bothered by symptoms. Patient perception also depends on their cultural and religious background.

Treatment satisfaction and other patient-reported outcomes are increasingly recognised as important measures of the effectiveness of care delivery in addition to clinical parameters.5 These measures allow health care providers to assess the quality of care that takes account of patient expectations and preferences. As more treatment options have become available, it would be of interest to assess this dimension of health care quality in patients with LUTS.

The aim of this survey was to explore South-East Asian patients’ perception of LUTS and their treatment. Self-reported prevalence, bother, treatment, and treatment satisfaction of selected symptoms including storage (urgency, nocturia), voiding (slow stream), and post-micturition symptoms (dribble) were evaluated. In addition, patients’ perception of the cause of nocturia and use of non-prescribed treatment for this symptom were explored. The study adopted a cross-sectional design and a convenience sample of patients who sought specialist care from across the region, namely Hong Kong, Malaysia, Philippines, Singapore, Thailand, and Vietnam. The study explored current unmet needs and will be used to guide the development of patient-centred management strategies. This sub-analysis evaluated data from the Hong Kong subpopulation.

We performed a cross-sectional study in 15 urology clinics—three from Hong Kong and 12 from South-East Asian countries (2 from Malaysia, 2 from the Philippines, 1 from Singapore, 3 from Thailand, and 4 from Vietnam). Hong Kong data were derived from a survey of patients conducted during December 2014 to October 2015. The study involved consecutive new patients who fulfilled the following eligibility criteria: (1) male aged ≥18 years who were attending the clinic for the first time, and (2) sought consultation at a urology clinic because of LUTS that had been present for at least 1 month. In this study, LUTS included increased daytime frequency, nocturia, urinary urgency, urinary incontinence, slow or weak stream, hesitancy, intermittency, straining, terminal dribble, sensation of incomplete bladder emptying, and post-micturition dribble. Such symptoms were defined according to the report from the Standardisation Sub-committee of the International Continence Society.6 Exclusion criteria were: (1) intermittent/indwelling catheterisation; (2) known prostate or bladder cancer; (3) spinal cord injury; (4) urethral stricture; (5) LUTS due to other causes, such as suspected or known urinary infection (cystitis or prostatitis); and (6) difficulty in understanding written information.

This survey involved major urology centres in South-East Asia. These centres were chosen because of two reasons: (1) they were major centres that saw 50 to 500 patients per week, and (2) there was a prior working relationship with the study group. Ethics approval from the Research Ethics Committee and Institutional Review Board were sought as required by the participating centres.

A patient self-administered questionnaire was developed by the Southeast Asia Urology Think Tank that comprised a group of urology specialists. The questionnaire assessed (a) demographics (country of residence, age-group); (b) presence/absence, bother, treatment, and treatment satisfaction of each of the following symptoms: urgency, nocturia, slow stream, and post-micturition dribble; (c) perception of the cause of disease and use of non-prescribed treatment for nocturia; and (d) where appropriate, satisfaction with transurethral resection of the prostate (TURP). The questionnaire in Chinese used in Hong Kong and the original questionnaire in English are shown in the Appendices. For questions that related to the level of ‘bother’, responses ranged from “bother me a lot”, “bother me some”, “bother me a little”, to “not at all bothersome”. For treatment satisfaction, patients were asked to choose from the four categories: “very satisfied”, “satisfied”, “unsatisfied”, and “very unsatisfied”.

The questionnaire was translated from the original English language into traditional Chinese. The translation of the questionnaire was performed in the ethnographic mode to maintain the meaning and cultural content. The questionnaire was back translated to the original language by an independent translator for each language. The versions were compared and any major differences were reconciled. An expert in survey research and a subject expert (ie a urologist) reviewed the translations to ensure the text was accurate, equivalent to the original, written at an appropriate level (ie understood by the general public), and took account of any cultural differences. A pre-test was performed in a sample of subjects prior to distribution to the entire sample to identify and correct any issues with clarity, relevance, and comprehensiveness, as well as user-friendliness of the questionnaire. The translations were revised after the pre-test and finalised by the urologist.

A urologist or member of the clinic staff was required to screen all consecutive new patients who presented to the urology clinic and identify potential study participants. They approached those patients who appeared eligible and explained the study to them. A patient questionnaire with an information sheet was distributed to each eligible consenting patient. Upon completion of the survey, the respondents sealed the questionnaire in an envelope and submitted it to the urologist who completed the eligibility checklist printed at the back of the questionnaire. The urologist then placed the questionnaire in an onsite lock box. Clinic staff periodically emptied the lock box and posted the contents to the data collection centre. Written informed consent was obtained from all eligible participants. To evaluate the response rate, participating centres were required to return a blank questionnaire for each eligible patient who refused to participate.

The sample size of 200 evaluable responses from each country was powered to detect a prevalence rate of 46% based on a previous study,6 with type 1 error of 0.05 and margin error not exceeding 7%. We used the SPSS (Windows version 24.0; IBM Corp, Armonk [NY], US) to analyse the data. Descriptive statistics were used to calculate the prevalence of LUTS, and the distribution of participants’ individual characteristics such as ‘bothersome level’, use of prescribed treatment, and satisfaction with specific treatments. Means and 95% confidence intervals are reported. Categorical variables are presented as percentages and compared using Chi squared test. Association rule analysis (also known as market basket analysis) was conducted using the statistical package R to examine symptom combination. Cochran-Armitage trend test was used to test the trend for nocturnal voiding frequency and the level of symptom-associated bother.

Between March and May 2014, a pilot survey was carried out to evaluate the applicability of the questionnaire and data collection procedure. The field test involved centres in Hong Kong, Malaysia, Philippines, Singapore, Thailand, and Vietnam. In each of these regions, the questionnaire was self-administered by approximately 30 respondents. A total of 233 questionnaires were received from all regions: high response rates of >90% were reported by the centres. Overall, respondents rated the questionnaire positively. The mean time taken to complete the questionnaire was 5.4 minutes.

Overall, clarity of the layout and instructions required improvement to minimise illogical/missing responses, for example, use of strong visual symbols such as arrows to direct respondents to the next question if the answer was YES/NO.

Of the 1436 questionnaires collected from all regions, 17 were deemed ineligible based on the exclusion criteria and three were blank. A total of 225 evaluable responses from Hong Kong were collated for analysis. Between December 2014 and October 2015, a total of 71 evaluable responses were collected from Prince of Wales Hospital, 65 from Queen Elizabeth Hospital and 89 from Tuen Mun Hospital.

In the Hong Kong sample, most respondents were aged 56 to 75 years (71%), retired (54%), and had not undergone TURP (80%). Compared with the non–Hong Kong population, Hong Kong respondents were older and more likely to have received TURP (P<0.05; Table 1).

Among the respondents in Hong Kong who sought care at the urology clinics, LUTS were prevalent, with nocturia the most common (93%; 95% confidence interval [CI], 90%-97%), followed by slow stream (76%; 95% CI, 71%-82%), post-micturition dribble (70%; 95% CI, 64%-76%), and urgency (50%; 95% CI, 43%-56%). Relative to the non–Hong Kong respondents, the prevalence was significantly higher for all four LUTS except urgency (Table 2).

A considerable number of respondents in Hong Kong reported having more than one symptom to some degree: 35% of them had all four symptoms, 32% reported three symptoms, 22% two symptoms, and only 9% a single symptom. With regard to symptom combinations, the strongest association was found for post-micturition dribble and co-existing urgency, nocturia, and slow stream (support: 0.353; confidence: 0.872; lift value: 1.249).

When asked to evaluate how bothersome their symptoms were, more than half of respondents in Hong Kong felt at least some degree of bother (ie “bother me some” or “bother me a lot”)—73% for urgency and post-micturition dribble, 70% for nocturia, and 68% for slow stream. When nocturia was defined as waking up twice or more at night, it became the symptom with which the most respondents reported at least some bother (77%). The Hong Kong respondents showed a consistent trend for a higher percentage for all symptoms studied compared with the non–Hong Kong group (Table 3).

Among the respondents in Hong Kong, 39% to 54% had received prescribed treatment. Less respondents with nocturia in Hong Kong received prescribed treatment compared with the non–Hong Kong population. The same applied to those with post-micturition dribble and urgency (Table 3).

More than 60% of respondents in Hong Kong reported being “satisfied” or “very satisfied” with previous prescribed treatment. Among the symptoms, the overall satisfaction rate was lower for nocturia and slow stream compared with the non–Hong Kong population. Most respondents in Hong Kong (79%-85%) indicated they would like to receive further treatment (Table 3).

A comparison of respondents’ subjective assessment of symptom bother, and outcomes of previous treatment from all regions is shown in Table 4.

Of those respondents who claimed to have nocturia, 77% reported at least two voids per night (Table 3). More frequent voiding was associated with a higher level of bother, ie more respondents reported “bother me some” or “bother me a lot” (Cochran-Armitage trend test, Z= –5.3044, P<0.0001; Fig). More than half (53%) of respondents regarded the prostate as the main cause of nocturia, and 39% answered “don’t know”. Some respondents (17%) had taken non-prescribed treatment for nocturia, including herbal medicines and supplements.

Only a small proportion (15%) of respondents had previously undergone TURP. More than half of respondents reported improved symptoms following surgery (Table 5). The small number of respondents and occurrence of multiple symptoms in individual respondents made it difficult to compare the improvement rates across symptoms.

The survey provided information on the characteristics of men with LUTS who attended urology clinics in Hong Kong. The respondents were mostly elderly, retired, and had not undergone any prostate surgery. The most commonly reported symptoms were nocturia and slow stream. Almost 80% of respondents with nocturia awoke at least twice per night to void. This was by far the most bothersome symptom followed by daytime symptoms, urgency, and post-micturition dribble. This finding is in line with results from previous studies conducted in a primary care setting that reported nocturia as a significant driving factor to trigger medical consultations.7 8 Nocturia could be highly bothersome to patients. Waking during the night to void disrupts sleep for both the patient and their partner, and impaired sleep is known to affect quality of life. We also established that LUTS often co-exist, and are in line with the EpiLUTS in Korea9 and Europe.10 In this study, the association rules analysis further revealed that in the local population, for respondents with urgency, slow stream and nocturia, there was an 87% risk of co-existing post-micturition dribble.

The survey findings help identify patient needs by exploring respondents’ perceived degree of bother and treatment satisfaction. Compared with the non–Hong Kong population, Hong Kong respondents tended to report a higher level of symptom bother (approximately 70% experiencing “some” or “a lot” bother), were less likely to have received prescribed treatment before consulting the current urologist, and reported lower treatment satisfaction. This could be due to several reasons: (1) cultural differences in patient perception of the symptom bother and their treatment expectation; (2) higher referral threshold for primary care practitioners in Hong Kong; (3) more conservative practice of primary care physicians (eg more use of watchful waiting); and (4) a longer waiting time to access urology specialty services.

One distinguishing feature of the Hong Kong health care system from other participating countries is that a primary care physician referral system is in place in Hong Kong. A greater involvement of the primary care sector in the management of LUTS is believed to allow a more rational use of resources as a result of reduced time spent on waiting lists, earlier therapy, and increased access to specialist services for those in need of extensive investigations or surgery.11 In this sense, it is understandable that it is those mostly highly symptomatic patients who have failed medical therapy are referred to urologists. Intriguingly, fewer respondents in Hong Kong than in other South-East Asian countries reported being given prescribed treatment, suggesting that primary care physicians adopt a more conservative approach to the treatment of LUTS, although this remains to be confirmed with findings from medical records.

Local data are limited whereas overseas studies showed that primary care physicians are more likely to engage in watchful waiting than urologists. Fewer men received recommended diagnostic evaluations from primary care physicians than from urologists, such as uroflowmetry, bladder scans, and cystoscopy. Primary care physicians ordered tests such as creatinine measurement that are less frequently used by urologists. Primary care physicians were also more likely to prescribe an older, long-acting alpha blocker that requires dose titration, and less likely to prescribe combination therapy such as 5-alpha-reductase inhibitor and anticholinergic therapy.12 13 Interdisciplinary care plans, detailed inter-physician referral and consultation letters, and integrated clinics have been suggested to improve physician communication and patient care.14 15

Satisfaction with care is an important outcome. It is multidimensional, comprising satisfaction with the care process and with care outcomes, and involves the patient’s personality and expectations and the health care provider’s interpersonal skills as well as technical excellence. The components of the care process may include waiting times, provision of information, access to care, adequacy of care environment, and speed of treatment. The other major element of the satisfaction domain concerns care outcome. Individuals whose outcome is either below personal expectations or who experience treatment side-effects may be less satisfied with the care they have received.16

In terms of the cause of nocturia, 39% of respondents responded “don’t know” and 53% regarded the prostate as the primary cause. This indicates a need for more education and better physician-patient communication. Surveys conducted in Korea and Europe showed that it is the complications of LUTS more than the actual symptoms with which respondents are most concerned.17 18 Because primary care physicians are usually the first point of medical contact for men with LUTS, they can play an important role in educating patients and diagnosing the condition.13

This is a large-scale multinational study that assesses LUTS and their treatment from the patients’ perspective at a secondary care level. Our study had several limitations. First, the subjects surveyed were patients who were bothered by their symptoms and who sought a urology opinion and agreed to participate in this study. Results of this self-administered survey were subject to recall bias, for examples, about medication received, number of voids at night, and quality of treatment received from their previous physicians. Further study is required to explore an association with clinical assessment. Second, convenience sampling limited the representativeness of the study sample. Third, the response rates of this study were estimated by the number of blank questionnaires returned; each of which indicated a refusal to participate. Centres might not follow this closely in practice and the reasons for non-participation were not recorded. This may have introduced potential response bias. Fourth, missing data are attributed to the involvement of older adults in this survey. Fifth, following the completion of the pilot survey, the questionnaire was later revised to improve clarity of the layout and instructions to minimise illogical/missing responses, for example, use of strong visual symbols such as arrows to direct respondents to the next question if the answer is YES/NO. Lastly, the questionnaires included only four predominant LUTS. Although these core symptoms are the main factors in the assessment of LUTS, it is possible that lack of assessment of other symptoms limited our understanding of all LUTS symptomatology. When interpreting the results, it is important to take into account the differences among cities/countries, referral systems, waiting list times, and socio-economic composition of the samples.

The demand for specialist care in the management of LUTS is expected to rise with increasing life-expectancy. Currently, TURP is the most common elective surgery in Hong Kong—about 3000 operations are performed at hospitals managed by the Hospital Authority each year.19 To keep pace with the increasing patient load and to improve the quality of care, findings from this study call for better streamlining of the shared-care model between general practitioners and urologists. More resources are required that will enable increased access to urology specialty services. This study provides an insight into the overall patient satisfaction with previous treatment before presentation to a specialist clinic. A complete evaluation of health care quality that examines different dimensions—such as structure, process, and quality of care—is warranted.

The survey provided much-needed information about patient perspectives regarding symptoms, bother, and treatment of LUTS. There is room for improvement in the quality of care provided to LUTS patients.

All authors have disclosed no conflicts of interest.

1. Yee CH, Li JK, Lam HC, Chan ES, Hou SS, Ng CF. The prevalence of lower urinary tract symptoms in a Chinese population, and the correlation with uroflowmetry and disease perception. Int Urol Nephrol 2014;46:703-10. Crossref

2. Wong SY, Woo J, Hong A, Leung JC, Kwok T, Leung PC. Risk factors for lower urinary tract symptoms in southern Chinese men. Urology 2006;68:1009-14. Crossref

3. Glover L, Gannon K, McLoughlin J, Emberton M. Men’s experiences of having lower urinary tract symptoms: factors relating to bother. BJU Int 2004;94:563-7. Crossref

4. Gannon K, Glover L, O’Neill M, Emberton M. Lower urinary tract symptoms in men: self-perceptions and the concept of bother. BJU Int 2005;96:823-7. Crossref

5. Abrams P, Andersson KE, Birder L, et al. Fourth International Consultation on Incontinence Recommendations of the International Scientific Committee: evaluation and treatment of urinary incontinence, pelvic organ prolapse, and fecal incontinence. Neurourol Urodyn 2010;29:213-40. Crossref

6. Li MK, Garcia LA, Rosen R. Lower urinary tract symptoms and male sexual dysfunction in Asia: a survey of ageing men from five Asian countries. BJU Int 2005;96:1339-54. Crossref

7. Abrams P, Cardozo L, Fall M, et al. The standardisation of terminology in lower urinary tract function: report from the standardisation sub-committee of the International Continence Society. Urology 2003;61:37-49. Crossref

8. Lai UC, Wun YT, Luo TC, Pang SM. In a free healthcare system, why do men not consult for lower urinary tract symptoms (LUTS)? Asia Pac Fam Med 2011;10:7. Crossref

9. Kim TH, Han DH, Lee KS. The prevalence of lower urinary tract symptoms in Korean men aged 40 years or older: a population-based survey. Int Neurourol J 2014;18:126-32. Crossref

10. Irwin DE, Milsom I, Hunskaar S, et al. Population-based survey of urinary incontinence, overactive bladder, and other lower urinary tract symptoms in five countries: results of the EPIC study. Eur Urol 2006;50:1306-14; discussion 1314-5. Crossref

11. Fonseca J, Martins da Silva C. The diagnosis and treatment of lower urinary tract symptoms due to benign prostatic hyperplasia by primary care family physicians in Portugal. Clin Drug Investig 2015;35(Suppl 1):19-27. Crossref

12. Spatafora S, Canepa G, Migliari R, et al. Effects of a shared protocol between urologists and general practitioners on referral patterns and initial diagnostic management of men with lower urinary tract symptoms in Italy: the Prostate Destination study. BJU Int 2005;95:563-70. Crossref

13. Miner MM. Primary care physician versus urologist: how does their medical management of LUTS associated with BPH differ? Curr Urol Rep 2009;10:254-60. Crossref

14. Lopéz BM, Romero AH, Ortín EO, Garcia IL. Can primary care physicians manage benign prostatic hyperplasia patients as urologists do? Eur Med J Urol Jul 2014:1-8.

15. Wei JT, Miner MM, Steers WD, et al. Benign prostate hyperplasia evaluation and management by urologists and primary care physicians: practice patterns from the observational BPH Registry. J Urol 2011;186:971-6. Crossref

16. Roehrborn CG, Nuckolls JG, Wei JT, Steers W; BPH Registry and Patient Survey Steering Committee. The benign prostatic hyperplasia registry and patient survey: study design, methods and patient baseline characteristics. BJU Int 2007;100:813-9. Crossref

17. George AK, Sandra MG. Measuring patient satisfaction. In: Penson DF, Wei JT, editors. Clinical research methods for surgeons. New Jersey: Humana Press; 2006: 253-65.

18. Kim SI, Kang JY, Lee HW, Seong do H, Cho JS. A survey conducted on patients’ and urologists’ perceptions of benign prostatic hyperplasia. Urol Int 2011;86:278-83. Crossref

19. Leung KK, So HS. Evolution of care for patients undergoing transurethral resection of prostate. Available from: https://www3.ha.org.hk/haconvention/hac2014/proceedings/downloads/SPPE559.pdf. Accessed 16 May 2016.

The term ‘inborn errors of metabolism’ (IEM) was coined by Archibald Garrod more than 100 years ago.1 Such disorder is extremely heterogeneous in clinical presentation and causes disease by either accumulation of toxic intermediary metabolites or lack of essential metabolites. In the 1960s, Robert Guthrie invented a bacterial-inhibition assay based on dried blood spots (DBS) collected on filter paper cards to detect abnormal levels of phenylalanine in patients with phenylketonuria (PKU).2 Newborn screening for PKU, other IEM, and non-IEM conditions (eg congenital hypothyroidism) became more widespread in the subsequent two decades. In the 1990s, tandem mass spectrometry (MS/MS) for multiplex analysis of acylcarnitines and amino acids was applied in expanded screening of IEM in newborns.3 4 Despite 50 years having passed since the first PKU screening, there are still vast differences in the practice of newborn screening in different countries.5 6 In the United States, the first recommended uniform newborn screening panel was adopted in 2006.7 In the latest recommended uniform screening panel, 20 of 34 core conditions and 22 of 26 secondary conditions are IEM with abnormal metabolites detectable by MS/MS.8 Screening of PKU was started in the 1960s in Australia, New Zealand, and Japan.9 10 Other Asia Pacific countries such as Taiwan, the Philippines, and Korea all have adopted different expanded newborn screening panels since then.11 Shanghai is the first city in China to adopt expanded newborn screening, starting in 2003.12 In Singapore, an expanded newborn screening programme that covers more than 25 IEMs was started in 2006.13

In Hong Kong, the only territory-wide newborn screening programme is cord blood screening for glucose-6-phosphate dehydrogenase (G6PD) deficiency and congenital hypothyroidism run by the Clinical Genetics Service, Department of Health.14 A pilot study using an OPathPaed service model for expanded newborn screening in a regional public hospital in Hong Kong was conducted between July and November 2010.15 The Centre of Inborn Errors of Metabolism (CIEM) of the Chinese University of Hong Kong started a private expanded newborn metabolic screening programme in July 2013 with participants from multiple centres. This study describes the results and screening outcome of this newborn screening programme in the past 3 years.

The CIEM newborn screening programme offered opt-in screening for 34 aminoacidopathies, organic acidurias, and fatty acid oxidation disorders (Table). Daily pre-test education and counselling were done by doctors and nurses of the referring units. This process was assisted by pamphlets produced by the CIEM.16 Parents were asked to sign a consent form after the education and counselling session. Referring hospitals were instructed to perform a heel prick for newborn babies between 24 hours and 7 days after birth and spot a few drops of blood onto a filter paper card provided by the CIEM. Apart from basic demographic information such as date and time of birth, the date and time of the DBS collection, ethnicity, feeding methods, medications, and family history of IEM were also collected. The screening laboratory ran the MS/MS assay for IEM screening daily from Monday to Friday. Eleven amino acids, succinylacetone, free carnitine, and 30 acylcarnitines were analysed by the Neobase non-derivatized MSMS kit (PerkinElmer, Waltham [MS], US) on a Quattro Micro tandem quadrupole mass spectrometer (Waters, Milford [MS], US). In the initial phase, laboratory cut-offs at 1 and 99 percentiles for these 43 analytes were calculated using results from 200 healthy newborns. These cut-offs were then updated regularly as more normal data were accumulated. We also compared our cut-offs with the clinically validated cut-offs in the Region 4 Stork (R4S) MS/MS project. The R4S MS/MS project is a web-based application for laboratory quality improvement of newborn screening by MS/MS.17 Screen-positive results were classified as ‘uncertain’ if the abnormal analyte(s) was only mildly elevated or ‘positive’ if the abnormal analyte(s) was markedly elevated or the abnormal analyte pattern was highly suggestive of a specific IEM. Post-analytical tools in the R4S MS/MS project (https://clir.mayo.edu/) were also used to assist result interpretation.18 The final screening reports were authorised by a trained chemical pathologist and a professionally qualified scientist. The clinical team, which consisted of metabolic paediatricians and newborn screening nurses, was notified immediately for any positive screening result. For each neonate with a positive screen, a second DBS card was collected. Additional blood and urine were usually collected at the same time for confirmatory metabolic investigations (eg plasma amino acid and urine organic acid analysis). The exact course of action was determined on a case-by-case basis. The workflow and logistics arrangement of the CIEM screening programme are summarised in the Figure.

Before and soon after launching of the CIEM newborn screening programme, a series of seminars and briefing sessions were organised in order to boost the knowledge of general practitioners, nurses, and laboratory staff on newborn screening. Continuous support was also provided to all the referring doctors and hospitals, especially on proper collection of DBS and interpretation of abnormal screening results. The CIEM also organises on-going yearly training to midwives about newborn screening.

Data for the CIEM newborn screening programme between July 2013 and July 2016 were retrieved. Basic demographics of the screened newborns, collection details for the DBS cards, call-back rate, false-positive rate, clinical details and outcomes of the true-positive cases, and call-back logistics were reviewed. This study was done in accordance with the principles outlined in the Declaration of Helsinki.

From July 2013 to July 2016, a total of 30 488 local newborn babies were screened. The total number of births during the same period was estimated to be 186 216.19 Therefore, approximately 16% of all newborns born between July 2013 and July 2016 were screened. The CIEM received DBS cards from the nursery units of nine of 10 private and two of eight public hospitals, and two paediatrics clinics in Hong Kong. Over 95% of the screened babies were Chinese and 2.7% were Caucasians. More than 98.3% of the DBS cards were collected within 7 days of life. Approximately 81% of the screening results were available in 2 calendar days and over 98% were available in 4 calendar days. Further analysis showed that most DBS cards with a turnaround time longer than 4 calendar days were received just before long holidays (eg the Lunar New Year holiday in 2014 and 2015). This is a well-known potential pitfall of a newborn screening programme, as most screening laboratories do not operate 7 days a week. In view of this, the CIEM added two extra half-day services during the Lunar New Year holiday in February 2016 to reduce the chance of delayed diagnosis.

Thirty-nine neonates had positive screening results (four ‘positive’ and 35 ‘uncertain’) and were called back for repeated DBS with or without additional metabolic investigations. The call-back rate was 0.128% (39/30 448). Six neonates (patients 1 to 6) were subsequently confirmed to have IEM by biochemical and molecular genetic testing. One neonate (patient 7) was confirmed to have abnormal newborn screening results due to a defect in maternal carnitine uptake. Details of patients 1 to 7 are described below. The false-positive rate was 0.105% (32/30 448). Among the 32 false-positive results, 17 had low free carnitine concentrations (range, 3.8-8.0 µmol/L) with or without low long-chain acylcarnitines, which constituted the most common cause of false-positive results.

Two siblings from the same Caucasian family were confirmed to have medium-chain acyl-coenzyme A dehydrogenase (MCAD) deficiency. Patient 1 was a full-term baby girl born by vaginal delivery. The first DBS sample showed marked elevations of C8-carnitine at 7.49 µmol/L (cut-off, <0.22 µmol/L) and other medium-chain acylcarnitines. The diagnosis of MCAD deficiency was confirmed by mutation analysis of the ACADM gene. The parents were counselled to feed their baby regularly and avoid fasting. Patient 1 was almost 3 years old at the time of the study and remained clinically asymptomatic. Patient 2 was the younger sister of patient 1. Her C8-carnitine concentration in the first DBS card collected before 48 hours after birth was 15.2 µmol/L. She shared the same ACADM genotype as patient 1 and also remained clinically well, and did not require active treatment.

Patient 3 was a boy with carnitine-acylcarnitine translocase (CACT) deficiency. He was born at 36 weeks and 2 days with a birth weight of 2.26 kg. The first DBS card was collected at 31 hours of life. He developed hypothermia, hypoglycaemia, hyperammonaemia, and bradycardia requiring active resuscitation with mechanical intubation at 42 hours of life. The newborn screening result was available at 50 hours of life and showed elevated C16-carnitine and C18:1-carnitine at 12.02 µmol/L (cut-off, <6.66 µmol/L) and 6.32 µmol/L (cut-off, <3.30 µmol/L), respectively. This acylcarnitine pattern was highly suggestive of CACT deficiency or carnitine palmitoyltransferase II deficiency. Follow-up genetic testing confirmed the diagnosis of CACT deficiency. Such deficiency is notorious for its early presentation in the postnatal period, with a high neonatal mortality rate.20 21 Although the newborn screening result was only available after patient 3 became symptomatic, early availability of the screening result has greatly assisted the neonatologists and metabolic paediatricians by guiding the direction of clinical management. With appropriate dietary and other management, the clinical condition of the patient remained good and he had normal neurodevelopment at 1.5 years of age.

Patient 4 was a girl with hyperphenylalaninaemia. She was born at 40 weeks of gestation with a birth weight of 3.45 kg. The first and second DBS showed elevated phenylalanine at 152 µmol/L and 89 µmol/L (cut-off, <88 µmol/L), respectively. Her urinary pterins were normal. Her plasma phenylalanine levels were monitored for 2 years, and ranged from 210 to 434 µmol/L while the patient was receiving an unrestricted diet. Genetic testing by sequence analysis and multiplex ligation–dependent probe amplification only detected a single mutation in the PAH gene. The patient received no specific dietary management and she had normal neurodevelopment up to the age of 2 years.

Patient 5 was a full-term boy with elevated C5-carnitine at 0.52 µmol/L (cut-off, <0.48 µmol/L) in his first newborn screen. Repeated DBS showed persistent elevation of C5-carnitine at 0.68 µmol/L. Urinary organic acid analysis showed elevated 2-methylbutylglycine. This result was highly suggestive of 2-methylbutyrylglycinuria (2-MBG) and excluded the more severe organic acid disorder isovaleric aciduria (IVA), which shared the same acylcarnitine marker as 2-MBG in newborn screening. The diagnosis was subsequently confirmed by genetic analysis of the ACADSB gene. 2-Methylbutylglycinuria is a relatively benign IEM and the patient had normal development at the age of 2 years without any specific treatment.

Patient 6 was a full-term girl whose first and second DBS showed elevated methionine at 138 µmol/L and 309 µmol/L (cut-off, <39 µmol/L), respectively. Following exclusion of homocystinuria, she was diagnosed with methionine adenosyltransferase deficiency by genetic testing. The patient was asymptomatic during the first year of life and was followed up at a metabolic clinic.

A mother with carnitine uptake defect (CUD) was diagnosed incidentally through the abnormal newborn screening result for her baby. The baby of this CUD patient had a low free carnitine level (2.1 ?mol/L; cut-off, >6.4 µmol/L) in the first DBS sample, which returned to normal on subsequent monitoring without the need for carnitine supplementation. Analysis of the mother showed that her serum free carnitine level was 1.08 µmol/L only. Maternal CUD causing low free carnitine in newborn screening was suspected. This was later confirmed by genetic analysis of the SLC22A5 gene. The mother had good past health throughout her life and during the pregnancy period. Her cardiac function was normal at the time of diagnosis. She was subsequently referred to a cardiologist for follow-up and was given carnitine supplementation therapy.

The CIEM screening programme did not have a mechanism to track false-negative results. Still one false-negative result was identified. The patient was a full-term baby girl with a body weight of 2.5 kg. She had newborn screening done on day 3 of life and the result was normal. In particular, her citrulline level was 19 µmol/L (cut-off, <30 µmol/L). She presented with prolonged jaundice at 1 month of age and was found to have a raised plasma citrulline level at 497 µmol/L (reference range, 3-35 µmol/L). Citrin deficiency was later confirmed by genetic analysis of the SLC25A13 gene.

This programme involved the participation of maternity units of 11 hospitals, including two public and nine private hospitals, and two paediatrics clinics. All 39 babies with positive screening results were successfully called back within an appropriate time-frame for follow-up investigations. Some of the call-backs were done by the referring doctors at their private clinics or hospitals while some were arranged by and done at the CIEM. Our experience showed that with proper education and professional support, the referring paediatricians were able to handle the call-backs of borderline abnormal results (eg free carnitine slightly below the cut-off) at the referring site and liaise with the CIEM for follow-up investigations. This practice not only lowered the workload of our metabolic paediatricians and newborn screening nurses, but also increased engagement of community paediatricians. For abnormal results requiring urgent clinical attention, the families were informed directly by our metabolic paediatricians, who would arrange urgent admission.

Expanded newborn screening for IEM by MS/MS has been widely adopted by many countries in the world for many years. Through early diagnosis and treatment, acute metabolic decompensations and long-term morbidity and mortality of many IEM can be prevented. In Hong Kong, medical practitioners and the general public are familiar with the cord blood screening programme for G6PD deficiency and congenital hypothyroidism, which is a very successful screening programme with highly satisfactory population coverage and outcome. However, little attention has been paid to IEM screening until recently.15 The Department of Health and Hospital Authority have done a pilot study on newborn screening for IEM in two public hospitals since October 2015. The establishment of the CIEM and its expanded newborn screening programme in 2013 has made this kind of screening service more readily accessible to local parents who understand the significance of IEM and opt for a private screening service for their newborn babies. The CIEM has successfully established a comprehensive screening programme, which comprises education, counselling, DBS collection, MS/MS screening, reporting, call-back, confirmatory investigations, and long-term follow-up and treatment. The CIEM screening programme quickly gained acceptance from private and public medical practitioners and now receives DBS cards from 11 hospital nursery units and two paediatrics clinics.

From July 2013 to July 2016, a total of 39 newborn babies were called back for abnormal screening results. Our experience showed that parental acceptance of abnormal screening results was generally good and this was likely the result of proper education and counselling before DBS collection. During the same period, we confirmed six cases of IEM through newborn screening and one false-negative case was identified. The incidence of IEM detected by this screening programme was one in 4355. The figure is very similar to that from previous local studies and other IEM prevalence studies in the Chinese population.11 12 22 23

Patient 3, with CACT deficiency, presented with hypoglycaemia and bradycardia before the newborn screening result was available. This is not unexpected because CACT deficiency is notorious for its early neonatal onset and high mortality rate.20 21 For this particular case, although newborn screening did not prevent the development of life-threatening clinical symptoms, it did provide an early diagnosis, which was extremely useful to the paediatricians. This case also illustrates that screening laboratories operating on a 5-day week may not be sufficient to meet the clinical needs and may delay the diagnosis of neonates born before or during weekends or long holidays. Operating a screening laboratory on a 7-day week, however, will increase the cost of the screening programme. A balance between the two is necessary.

By screening the abnormal analytes for important IEM, some less clinically important IEM may be revealed. For example, patient 5 had elevated C5-carnitine, which has two main differential diagnosis, one is IVA and the other is 2-MBG. Of note, IVA is an important organic acid disorder. Affected patients usually present with hyperammonaemia, metabolic acidosis, and acute metabolic decompensation. On the other hand, 2-MBG is a disorder of uncertain clinical significance. Although the exact clinical course is not yet clear, there is no case report demonstrating a definite clinical correlation between 2-MBG and any long-term mortality and morbidity. After conducting a review in February 2016, the CIEM decided to remove 2-MBG from the list of target IEM. This can minimise the potential harm of labelling an otherwise healthy neonate with a life-long label of an IEM of uncertain clinical significance that does not require any treatment.

Until a government-funded universal expanded newborn screening service is available in Hong Kong, private medical practitioners are charged with the task of selecting newborn screening service providers for their clients. Some medical practitioners may focus on the number of screening targets when they choose a screening service provider and mistakenly believe the more the better. Nonetheless, we should not ignore the harm (eg the anxiety generated by a false-positive result) of over-screening. Other than the appropriateness of the screening targets, the turnaround time of the screening tests and the availability of confirmatory investigations are also crucial. Many screening targets may present in the early neonatal period. For a screening test to exert its maximum benefits, the results should be available within a reasonable time-frame. No newborn screening tests are confirmatory by themselves. Therefore, whenever there is a positive newborn screening result, further investigations to confirm or refute the diagnosis must be in place. Medical practitioners who use a private newborn screening service must be aware of this point and ensure all further investigations generated by a positive newborn screening result are acceptable and affordable by their patients. The use of spot urinary specimens instead of DBS for newborn screening is appealing to parents as collection of urine does not require a heel prick and thus appears to be non-invasive. Parents should be educated that heel prick using standard devices and done by well-trained nurses or phlebotomists are non-traumatic and generate no harm to newborn babies. They should also be made aware that DBS is the standard sample of choice adopted by most, if not all, national newborn screening programmes.

The scale and duration of this study is far from sufficient to draw any conclusion on financial benefits or cost-effectiveness of expanded newborn screening. From a public service perspective, a condition is appropriate for screening if early diagnosis has demonstrated benefits and is cost-effective. Both local and international studies have shown the cost-efficiencies gained by adopting MS/MS technology for expanded newborn screening.24 25 26 The available evidence is sufficient for policymakers to consider implementing a universal expanded screening programme in Hong Kong.

The CIEM has established a comprehensive expanded newborn screening programme for selected IEM. The programme involves pre-test counselling, a good logistics arrangement for efficient incoming referral and reporting, a readily available confirmatory testing service and, most importantly, timely management by medical and nursing specialists. Each component contributes towards a successful newborn screening programme. This screening programme not only increases the awareness of local health care workers and the general public of newborn screening, but also sets a standard against which the performance of other private newborn screening tests can be compared. In the Hong Kong SAR Chief Executive’s Policy Address 2017, it was announced that the government plans to extend its pilot newborn screening service from two to all public hospitals with maternity wards in phases from the second half of 2017-18.27 Our experience could serve as a reference for policymakers when they contemplate establishing a government-funded universal expanded newborn screening programme in the future.

The CIEM would like to express sincere gratitude to the Joshua Hellmann Foundation for Orphan Diseases for their generous donation and continuous support. The Foundation had no role in the design of the study; collection, analysis, or interpretation of the data; writing, review, or approval of the manuscript; or the decision to submit the manuscript for publication.

The authors have disclosed no conflicts of interest.

1. Garrod AE. The incidence of alkaptonuria: a study in chemical individuality. Lancet 1902;160:1616-20. Crossref

2. Guthrie R, Susi A. A simple phenylalanine method for detecting phenylketonuria in large populations of newborn infants. Pediatrics 1963;32:338-43.

3. Millington DS, Norwood DL, Kodo N, Roe CR, Inoue F. Application of fast atom bombardment with tandem mass spectrometry and liquid chromatography/mass spectrometry to the analysis of acylcarnitines in human urine, blood, and tissue. Anal Biochem 1989;180:331-9. Crossref

4. Chace DH, Millington DS, Terada N, Kahler SG, Roe CR, Hofman LF. Rapid diagnosis of phenylketonuria by quantitative analysis for phenylalanine and tyrosine in neonatal blood spots by tandem mass spectrometry. Clin Chem 1993;39:66-71.

5. Boyle CA, Bocchini JA Jr, Kelly J. Reflections on 50 years of newborn screening. Pediatrics 2014;133:961-3. Crossref

6. Therrell BL, Padilla CD, Loeber JG, et al. Current status of newborn screening worldwide: 2015. Semin Perinatol 2015;39:171-87. Crossref

7. American College of Medical Genetics Newborn Screening Expert Group. Newborn screening: toward a uniform screening panel and system—executive summary. Pediatrics 2006;117(5 Pt 2):S296-307.

8. Advisory Committee on Heritable Disorders in Newborns and Children. Recommended uniform screening panel. Available from: https://www.hrsa.gov/advisorycommittees/mchbadvisory/heritabledisorders/recommendedpanel/. Accessed 17 Jan 2017.

9. Padilla CD, Therrell BL. Newborn screening in the Asia Pacific region. J Inherit Metab Dis 2007;30:490-506. Crossref

10. Tada K, Tateda H, Arashima S, et al. Follow-up study of a nation-wide neonatal metabolic screening program in Japan. A collaborative study group of neonatal screening for inborn errors of metabolism in Japan. Eur J Pediatr 1984;142:204-7. Crossref

11. Niu DM, Chien YH, Chiang CC, et al. Nationwide survey of extended newborn screening by tandem mass spectrometry in Taiwan. J Inherit Metab Dis 2010;33(Suppl 2):S295-305. Crossref

12. Gu X, Wang Z, Ye J, Han L, Qiu W. Newborn screening in China: phenylketonuria, congenital hypothyroidism and expanded screening. Ann Acad Med Singapore 2008;37(12 Suppl):107-4.

13. Lim JS, Tan ES, John CM, et al. Inborn Error of Metabolism (IEM) screening in Singapore by electrospray ionization-tandem mass spectrometry (ESI/MS/MS): An 8 year journey from pilot to current program. Mol Genet Metabo 2014;113:53-61. Crossref

14. Lam ST, Cheng ML. Neonatal screening in Hong Kong and Macau. Southeast Asian J Trop Med Public Health 2003;34 Suppl 3:73-5.

15. Mak CM, Lam C, Siu W, et al. OPathPaed service model for expanded newborn screening in Hong Kong SAR, China. Br J Biomed Sci 2013;70:84-8.

16. Joshua Hellmann Foundation–Newborn metabolic screening program. Available from: http://www.obg.cuhk.edu.hk/fetal-medicine/fetal-medicine_services/iem/. Accessed Jan 2017.

17. McHugh D, Cameron CA, Abdenur JE, et al. Clinical validation of cutoff target ranges in newborn screening of metabolic disorders by tandem mass spectrometry: a worldwide collaborative project. Genet Med 2011;13:230-54. Crossref

18. Hall PL, Marquardt G, McHugh DM, et al. Postanalytical tools improve performance of newborn screening by tandem mass spectrometry. Genet Med 2014;16:889-95. Crossref

19. Population Estimates. Census and Statistical Department, The Hong Kong SAR Government. Available from: https://www.censtatd.gov.hk/hkstat/sub/sp150.jsp?tableID=004&ID=0&productType=8. Accessed 28 Jun 2017.

20. Vitoria I, Martín-Hernández E, Peña-Quintana L, et al. Carnitine-acylcarnitine translocase deficiency: experience with four cases in Spain and review of the literature. JIMD Rep 2015;20:11-20. Crossref

21. Lee RS, Lam CW, Lai CK, et al. Carnitine-acylcarnitine translocase deficiency in three neonates presenting with rapid deterioration and cardiac arrest. Hong Kong Med J 2007;13:66-8.

22. Lee HC, Mak CM, Lam CW, et al. Analysis of inborn errors of metabolism: disease spectrum for expanded newborn screening in Hong Kong. Chin Med J (Engl) 2011;124:983-9.

23. Hui J, Tang NL, Li CK, et al. Inherited metabolic diseases in the Southern Chinese population: spectrum of diseases and estimated incidence from recurrent mutations. Pathology 2014;46:375-82. Crossref

24. Cipriano LE, Rupar CA, Zaric GS. The cost-effectiveness of expanding newborn screening for up to 21 inherited metabolic disorders using tandem mass spectrometry: results from a decision-analytic model. Value Health 2007;10:83-97. Crossref

25. Ng VH, Mak CM, Johnston JM. Cost-effectiveness analysis of newborn screening for organic acidemias in Hong Kong. SM J Clin Pathol 2016;1:1001.

26. Grosse SD, Thompson JD, Ding Y, Glass M. The use of economic evaluation to inform newborn screening policy decisions: the Washington State experience. Milbank Q 2016;94:366-91. Crossref

27. The 2017 Policy Address. Available from: https://www.policyaddress.gov.hk/2017/eng/pdf/PA2017.pdf. Accessed Aug 2017.

Eczema is a chronic inflammatory skin disease associated with cutaneous hyperreactivity to environmental stimuli such as microbial exposure and food ingestion that are otherwise tolerated by unaffected subjects.1 In our community study, nearly one third of preschool children had eczema and 8.1% had parent-reported adverse food reactions.2 Thus, there is a significant health care burden from both eczema and food allergy in Hong Kong. Food allergy is believed by many patients and their family to be a major cause of eczema. Many traditional Chinese medicine practitioners also advise extensive food avoidance for eczema. Our previous study reported that over half of local children with eczema practised food avoidance.3 There are, however, limited objective data on the intake of specific food items and nutrients in children with eczema. Henderson and Hayes4 reported the correlation between dietary calcium intake and bone mineral density (BMD) in 55 children and adolescents aged 5 to 14 years with cow’s milk sensitisation. Their calcium intake was determined using a food frequency questionnaire (FFQ). The bone density Z-score of their patients with a milk allergy serially increased with increasing calcium intake. In another study, Jensen et al5 investigated BMD in children aged 8 to 17 years with verified cow’s milk allergy (CMA) for more than 4 years and compared them with 343 healthy controls. Their patient’s BMD was markedly reduced for age, and height for age was reduced indicating ‘short’ bones. Furthermore, calcium consumption was about 25% of that recommended. In another study, BMD was assessed in 27 young children with CMA.6 During bone resorption, osteoclasts secrete a mixture of acid and neutral proteases that degrade the collagen fibrils into molecular fragments including C-terminal telopeptide. Its aspartic acid changes from alpha to beta form as bone ages. The latter form called beta-crosslaps is a specific marker for bone resorption; its concentration was lower in CMA patients than in controls, indicating an increased bone turnover in the former. Ten CMA patients had a BMD Z-score of less than –1 standard deviation (SD) value. Despite these results, there was limited evidence of compromised bone health in eczema patients when such results were adjusted for confounders such as calcium intake and physical activity. This study investigated intake of food items and nutrients as well as BMD in Chinese children in Hong Kong with varying degrees of eczema, and compared these values with unaffected children.

This cross-sectional study recruited ethnic Chinese children aged below 18 years with physician-diagnosed eczema from our paediatric allergy and dermatology clinics over a 6-month period in 2012. Eczema was diagnosed using standard criteria,7 and disease severity was assessed by SCORing Atopic Dermatitis.8 Patients were treated with topical mometasone furoate only. Those prescribed other topical steroids were changed to this drug for at least 4 weeks before the study. Patients prescribed oral immunosuppressive drugs within 6 months were excluded. Subjects with stable asthma and/or allergic rhinitis who were free of eczema and food allergy as well as non-allergic children were recruited from attendants at our out-patient clinics as a reference group. As an inclusion criterion, subjects in the reference group had no dietary restrictions. Following informed written consent, our research staff recorded subjects’ intake of a wide spectrum of dietary components using a Chinese FFQ. Children who attended secondary school and beyond were assessed using the adolescent/adult version,9 10 whereas those in primary schools and below were assessed using the preschool version.11 Intake of individual food items and groups of major nutrients as recorded in FFQ were quantified and analysed by Foodworks Professional software (version 7; Xyris, Brisbane, Australia). This software included only certain Chinese food items commonly found in Australia. Thus, our co-author (FYYK), who has a dietetic qualification, added new Chinese food data with reference to a comprehensive Chinese food composition database published in 2004 by the Peking University Medical Press (available upon request). These new food items were saved in the software as a new food composition database so that their nutrient data could be computed. Physical activity level of participants was assessed using a Physical Self-Description Questionnaire.12 The study was approved by the Clinical Research Ethics Committee of our university. All participants provided written informed consent.

Urinary concentrations of cross-linked N-telopeptides of type 1 collagen (NTx), biomarkers of bone resorption,13 were measured by enzyme-linked immunosorbent assay (Wampole Laboratories, Princeton [NJ], US) with a lower limit of detection set at 20 nM of bone collagen equivalent. Results were corrected for creatinine that was measured by modified Jaffe reaction (Roche Diagnostics GmbH, Mannheim, Germany).

Participants’ BMD was measured at the mid-point of the radius in the non-dominant arm and at the left tibia using quantitative ultrasound bone sonometry (QUBS) to determine the velocity of ultrasound wave, expressed as the speed of sound in m/s, using Omnisense 7000P (BeamMed, Petach Tikva, Israel) as described previously.14 15 This machine compared speed of sound measurements to a built-in reference database of a healthy urban Chinese population of 797 boys and 760 girls aged 0 to 18 years. As described in the manufacturer’s manual, subjects with previous osteoporotic fractures, use of medications affecting bone health, presence of a disease known to affect bone metabolism, recent prolonged immobilisation, or a systemic malignant disease within 5 years were excluded from such reference database. The Omnisense devices were transported from place to place, and the same group of operators performed all measurements. Results of BMD were then expressed as age- and gender-matched Z-score.

Dietary food and nutrient intake, BMD Z-scores, and urine NTx levels between different groups were analysed by Student’s t test or analysis of variance (ANOVA). The relationship between eczema and dietary, BMD, and urinary variables that differed significantly between children with eczema and those in the reference group were confirmed by multivariable stepwise binary logistic regression adjusted for age, gender, body mass index (BMI), physical activity level, and co-morbid allergic diseases as covariates. The correlations between clinical variables, BMD Z-scores, and urine NTx levels were analysed by Pearson correlation. Continuous variables with skewed data distribution were transformed to achieve normal distribution prior to analyses. All analyses were performed with the use of SPSS (Windows version 18.0; SPSS Inc, Chicago [IL], US). The level of significance was set at 0.05, and all P values were two-tailed.

A total of 114 children with eczema and other 60 children as the reference group were recruited (Table 1). The means (± SDs) age of children in the reference group, children with mild eczema, and children with moderate-to-severe eczema were 10.0 ± 5.0, 9.8 ± 5.4, and 11.9 ± 3.8 years, respectively. Age, gender, and anthropometric variables did not differ among children with eczema and those in the reference group. Table 2 describes the children’s daily food intake adjusted for total energy as recorded by FFQ. Males had a higher daily energy intake than females (median [interquartile range]: 7570 [6267-9487] kJ vs 6736 [5438-8547] kJ; P=0.035), but intake of any single food item or nutrient did not differ when adjusted for daily energy intake. Multivariable stepwise binary logistic regression analyses revealed that a higher intake of soybeans and soybean products as well as miscellaneous dairy products was associated with an increased risk of eczema, although statistically significant associations were only found between the third tertile of soybean intake and mild eczema, as well as between the second tertile and mild eczema, and between the third tertile and moderate-to-severe eczema for intake of miscellaneous dairy products. A higher consumption of eggs, beef, and shellfish was associated with lower risk of eczema (Table 3). Children with eczema and those in the reference group consumed similar supplements of cod liver oil, fish oil, vitamins, and calcium (P>0.9 for all; data not shown). Table 4 summarises their nutrient intake. Patients with moderate-to-severe eczema consumed a lower amount of vitamin D, calcium, and iron when compared with those with mild eczema and/or those in the reference group. Children with the highest tertile of intake for vitamin D (odds ratio [OR]=0.16; 95% confidence interval [CI] 0.05-0.48; P=0.001) and calcium (OR=0.17; 95% CI, 0.06-0.51; P=0.002) had a lower risk for moderate-to-severe eczema than those with the lowest tertile when adjusted for age, gender, BMI Z-score, and physical activity level as covariates.

The mean BMD Z-score for children with eczema and those in the reference group was 0.52 and 0.55 over the radius (P=0.889), and 0.02 and –0.01 over the tibia (P=0.886) [Table 5]. The BMD did not differ between children with mild and moderate-to-severe eczema over these two regions (P=0.296 and 0.661, respectively). The BMD Z-score at the tibia correlated inversely with children’s age (r = –0.282, P<0.001), but the Z-score at both regions was independent of BMI Z-score. Age of onset of eczema did not influence BMD Z-score at the radius (P=0.349) or tibia (P=0.240), nor was it associated with physical activity level (P>0.1 for both). Urine NTx levels did not differ between patients and reference subjects (P=0.451; Table 5), between reference subjects and those with mild or moderate-to-severe eczema, or between those with mild and moderate-to-severe eczema (P>0.3 for all). This biomarker showed an inverse correlation with subject’s age (r = –0.569, P<0.001) but not BMI Z-score (r = –0.132, P=0.101) or BMD Z-score at the radius (r = –0.133, P=0.097) or tibia (r = –0.135, P=0.093). Repeated analyses in eczema or reference subject subgroups yielded similar results.

Regarding the possible effects of calcium intake, the tertiles of daily intake were not associated with BMD Z-score at either the radius or tibia among reference subjects (Table 6). On the other hand, patients at the second tertile of calcium intake had a lower BMD Z-score at both the radius and tibia than those at the first and third tertiles (P=0.016 and 0.015, respectively by one-way ANOVA). Patients with the highest tertile of calcium intake had higher BMD Z-score at the tibia.

This study is the first to report the effects of eczema and dietary intake on BMD in Chinese children in Hong Kong. Children with eczema had a higher intake of soybeans and miscellaneous dairy products but lower intake of eggs, beef, shellfish, vitamin D, calcium, and iron than the reference group. Despite these differences, children with eczema had similar BMD Z-scores at the radius and tibia and urinary NTx levels. We also observed a trend for patients with the highest tertile of calcium intake to have a higher BMD Z-score at the tibia.

Food allergy is commonly believed to be a major cause of eczema in the Chinese culture. Werfel and Breuer16 supported this by the finding that atopic dermatitis could be exacerbated by certain foods in more than 50% of affected children. Nonetheless, reactions induced by classic foods such as hen’s eggs and cow’s milk were less common in adolescents and adults than in young children. Food allergy in eczema may be immunoglobulin (Ig) E–mediated or non–IgE-mediated, thus food-induced eczema should be diagnosed only by thorough clinical history-taking and diagnostic work-up. Because of a possible co-existence of eczema and food allergy, dietary restrictions form an integral component of eczema management in children. A systematic review of 421 participants from nine randomised controlled trials only suggested some benefit of an egg-free diet in infants with suspected egg allergy who had positive specific IgE to eggs. No benefit, however, could be detected for the use of various exclusion diets in unselected people with atopic eczema.17 Clinicians should also bear in mind the dynamic nature of food allergy, and that young children might outgrow such allergies.

Eczema severity was associated with altered dietary intake in our children. Of Hong Kong children aged 2 to 7 years, 8% reported adverse food reactions, with the six leading foods being shellfish, eggs, peanuts, beef, cow’s milk, and tree nuts.2 Such adverse reactions to multiple foods also led to worse quality of life.18 In the EuroPrevall study, shrimp, mango, milk, eggs, and peanuts were the foods most commonly reported to cause allergy among primary schoolchildren in Hong Kong.19 20 According to the presence of allergen-specific IgE, however, milk and egg sensitisation was identified in over 14% of these subjects whereas that for all other foods including shrimp and peanuts was less than 8%. In our hospital-based patients, skin prick testing revealed that shellfish, peanuts, nuts, and eggs were the most common food allergens.21 In general, allergy to milk, eggs, beef, and seafood is a common cultural belief in Chinese families with children having eczema. Consistently, our patients with moderate-to-severe eczema had a lower intake of eggs, beef, and shellfish as well as vitamin D, calcium, and iron (Tables 2 3 4). To compensate for such dietary restrictions, they ingested more soybeans and soybean products, as well as miscellaneous dairy products. These alterations were consistent with our earlier findings that dietary restrictions to avoid high calcium foods such as cow’s milk were common among Hong Kong children with eczema.3 22 Milk intake was also negatively associated with eczema diagnosis among Spanish primary schoolchildren.23

Dietary intake of vitamin D was very low in our children, and virtually all (both cases and controls) ingested a lower amount than the age- and sex-specific dietary reference intake for the Chinese population (Table 4). Our data also suggested an inverse relationship between eczema severity and vitamin D intake. These results echoed those of our recent study in which low serum vitamin D level was highly prevalent in both Hong Kong children with eczema and non-allergic controls.24 Vitamin D deficiency was associated with disease severity in our eczema children. Our study found eczema severity to be associated with several single-nucleotide polymorphisms of vitamin D pathway genes.25 Besides its role in bone metabolism, vitamin D3 exerted pluripotent effects on both cutaneous adaptive (eg T-cell activation and dendritic cell maturation) and innate (eg expression of antimicrobial peptides) immunity.26 27 Our data suggested low serum levels of LL-37, the biologically active form of cathelicidin involved in antimicrobial defence, in children with eczema.28 Alterations in local vitamin D3 levels also modulated skin barrier function.29 Most children in Hong Kong, a subtropical region, had long school hours on weekdays and swam mainly in indoor pools. Thus, they are not expected to produce enough vitamin D from sunlight exposure. Together with our dietary data, we recommend that local children increase their outdoor activities and dietary intake of vitamin D.

Dietary intake of our children was recorded by FFQ. This method has been used to assess habitual intake over extended periods of time, ranging from the past month to the past year, by asking respondents to report frequency of consumption and often portion size for a defined list of foods and beverages.30 This allows for investigation of individual dietary pattern, ranking of usual individual intake, and examination of associations between frequency of consumption of certain items and individual clinical conditions. Woo et al9 developed an FFQ for the Hong Kong Chinese population that was later adapted and validated in local children and adolescents.11 12 13 18 Energy intake of two thirds of our subjects was below the Chinese-specific dietary reference intake (Table 4). Based on our data, FFQ would overestimate the intake of energy and macronutrients when compared with a 3-day food record.11 Thus, both our patients and controls had an inadequate nutritional intake.

Skeletal problems were common in children with food allergy and eczema. The BMD in children aged 8 to 17 years with CMA was markedly reduced for age, and calcium consumption was only one quarter that recommended.5 Beta-crosslaps concentration as a biomarker of bone turnover was lower in CMA patients than in controls.6 Osteoporosis and osteopenia were detected in 4.8% and 32.8% of 125 adults with moderate-to-severe eczema, respectively.31 Low BMD in these patients did not seem to respond to calcium and/or vitamin D supplementation.32 Another small adult study found similar BMD between subjects with eczema and controls.33 In childhood eczema, patients with severe disease treated with topical corticosteroids and cyclosporin had much lower BMD as measured by dual energy X-ray absorptiometry (DEXA) than those prescribed topical corticosteroids alone.34 The BMD was similar between Dutch children with moderate-to-severe eczema and the general population.35 Overall, there are limited data regarding BMD in childhood eczema. In this study, children with eczema had similar sonographically measured BMD at the radius and tibia and urinary level of NTx when compared with controls (Table 5). Nonetheless, BMD Z-scores were lower among patients in the second than third tertile of dietary calcium intake (Table 6). We do not recommend unjustified restriction of calcium-rich foods such as cow’s milk and soybean in children with eczema.

This study has several limitations. First, we did not record subjects’ outdoor activities or measure their serum vitamin D level. Second, this study assessed subjects’ dietary intake using a FFQ instead of a 3-day food record. The FFQ is cheap and easy to administer. Such FFQ with parental reporting was also a reasonably valid way to collect dietary data on children even in situations when parents do not observe all meals and snacks eaten by their child.36 37 38 Third, because of its cross-sectional nature, this study did not collect information about the duration of food avoidance. Fourth, BMD of our children was measured by ultrasound rather than DEXA; with the latter being the gold standard for diagnosing osteoporosis. There is substantial concern about radiation hazard especially in children.39 This study adopted the radiation-free technique of QUBS that was useful in assessing BMD in children.40 Lastly, our patients with moderate-to-severe eczema were nearly 2 years older than the reference group although patients as a whole group were of similar age (Table 1). Because of the higher energy needs of older and bigger children, we adjusted for subjects’ age and gender in multivariable regression analyses and compared their dietary intake with age- and gender-specific dietary reference intakes.

Dietary restrictions are common among Chinese children with eczema in Hong Kong. These patients had a lower calcium, vitamin D, and iron intake. Despite this, childhood eczema was not associated with diminished BMD. Nonetheless, a significant association was detected between calcium intake and BMD among these patients.

This work was funded by Direct Grants for Research (2011.1.058 and 2013.2.033) of the Chinese University of Hong Kong. We thank Yvonne YF Ho and Patty PP Tse for helping with patient assessment and data collection.

The authors have disclosed no conflicts of interest.

1. Leung DY, Bieber T. Atopic dermatitis. Lancet 2003;361:151-60. Crossref

2. Leung TF, Yung E, Wong YS, Lam CW, Wong GW. Parent-reported adverse food reactions in Hong Kong Chinese pre-schoolers: epidemiology, clinical spectrum and risk factors. Pediatr Allergy Immunol 2009;20:339-46. Crossref

3. Hon KL, Leung TF, Kam WY, Lam MC, Fok TF, Ng PC. Dietary restriction and supplementation in children with atopic eczema. Clin Exp Dermatol 2006;31:187-91. Crossref

4. Henderson RC, Hayes PR. Bone mineralization in children and adolescents with a milk allergy. Bone Miner 1994;27:1-12. Crossref

5. Jensen VB, Jørgensen IM, Rasmussen KB, Mølgaard C, Prahl P. Bone mineral status in children with cow milk allergy. Pediatr Allergy Immunol 2004;15:562-5. Crossref

6. Hidvégi E, Arató A, Cserháti E, Horváth C, Szabó A, Szabó A. Slight decrease in bone mineralization in cow milk-sensitive children. J Pediatr Gastroenterol Nutr 2003;36:44-9. Crossref

7. Hanifin JM, Rajka G. Diagnostic features of atopic dermatitis. Acta Derm (Stockh) 1980;92:44-7.

8. Hon KL, Leung TF, Wong Y, Fok TF. Lesson from performing SCORADs in children with atopic dermatitis: subjective symptoms do not correlate well with disease extent or intensity. Int J Dermatol 2006;45:728-30. Crossref

9.Woo J, Leung SS, Ho SC, Lam TH, Janus ED. A food frequency questionnaire for use in the Chinese population in Hong Kong: description and examination of validity. Nutr Res 1997;17:1633-41. Crossref

10. Chan RS, Woo J, Chan DC, Cheung CS, Lo DH. Estimated net endogenous acid production and intake of bone health-related nutrients in Hong Kong Chinese adolescents. Eur J Clin Nutr 2009;63:505-12. Crossref

11. Kwok FY, Ho YY, Chow CM, So CY, Leung TF. Assessment of nutrient intakes of picky-eating Chinese preschoolers using a modified food frequency questionnaire. World J Pediatr 2013;9:58-63. Crossref

12. Yu CC, Sung RY, Hau KT, Lam PK, Nelson EA, So RC. The effect of diet and strength training on obese children’s physical self-concept. J Sports Med Phys Fitness 2008;48:76-82.

13. Bollen AM, Eyre DR. Bone resorption rates in children monitored by the urinary assay of collagen type 1 cross-linked peptides. Bone 1994;15:31-4. Crossref

14. Jones G, Boon P. Which bone mass measures discriminate adolescents who have fractured from those who have not? Osteoporosis Int 2008;19:251-5. Crossref

15. Christoforidis A, Printza N, Gkogka C, et al. Comparative study of quantitative ultrasonography and dual-energy X-ray absorptiometry for evaluating renal osteodystrophy in children with chronic kidney disease. J Bone Miner Metab 2011;29:321-7. Crossref

16. Werfel T, Breuer K. Role of food allergy in atopic dermatitis. Curr Opin Allergy Clin Immunol 2004;4:379-85. Crossref

17. Bath-Hextall F, Delamere FM, Williams HC. Dietary exclusions for improving established atopic eczema in adults and children: systematic review. Allergy 2009;64:258-64. Crossref

18. Leung TF, Yung E, Wong YS, Li CY, Wong GW. Quality-of-life assessment in Chinese families with food-allergic children. Clin Exp Allergy 2009;39:890-6. Crossref

19. Wong GW, Mahesh PA, Ogorodova L, et al. The EuroPrevall-INCO surveys on the prevalence of food allergies in children from China, India and Russia: the study methodology. Allergy 2010;65:385-90. Crossref

20. Wong GW, Ogorodova L, Mahesh PA, et al. Food allergy in schoolchildren from China, Russia and India: the EuroPrevall-INCO surveys. Allergy 2011;66 Suppl 94:27.

21. Hon KL, Wang SS, Wong WL, Poon WK, Mak KY, Leung TF. Skin prick testing in atopic eczema: atopic to what and at what age? World J Pediatr 2012;8:164-8. Crossref

22. Hon KL, Leung TF, Lam MC, et al. Eczema exacerbation and food atopy beyond infancy: how should we advise Chinese parents about dietary history, eczema severity and skin prick testing? Adv Ther 2007;24:223-30.

23. Suárez-Varela MM, Alvarez LG, Kogan MD, et al. Diet and prevalence of atopic eczema in 6 to 7-year-old schoolchildren in Spain: ISAAC phase III. J Investig Allergol Clin Immunol 2010;20:469-75.

24. Wang SS, Hon KL, Kong AP, Pong HN, Wong GW, Leung TF. Vitamin D deficiency is associated with diagnosis and severity of childhood atopic dermatitis. Pediatr Allergy Immunol 2014;25:30-5. Crossref

25. Wang SS, Hon KL, Kong AP, et al. Eczema phenotypes are associated with multiple vitamin D pathway genes in Chinese children. Allergy 2014;69:118-24. Crossref

26. Bikle DD. What is new in vitamin D: 2006-2007. Curr Opin Rheumatol 2007;19:383-8. Crossref

27. Schauber J, Dorschner RA, Yamasaki K, Brouha B, Gallo RL. Control of the innate epithelial antimicrobial response is cell-type specific and dependent on relevant microenvironmental stimuli. Immunology 2006;118:509-19. Crossref

28. Leung TF, Ching KW, Kong AP, Wong GW, Chan JC, Hon KL. Circulating LL-37 is a biomarker for eczema severity in children. J Eur Acad Dermatol Venereol 2012;26:518-22. Crossref

29. Bikle DD, Chang S, Crumrine D, et al. Mice lacking 25OHD 1alpha-hydroxylase demonstrate decreased epidermal differentiation and barrier function. J Steroid Biochem Mol Biol 2004;89-90(1-5):347-53. Crossref

30. Thompson FE, Byers T. Dietary assessment resource manual. J Nutr 1994;124(11 Suppl):2245S-2317S.

31. Haeck IM, Hamdy NA, Timmer-de Mik L, et al. Low bone mineral density in adult patients with moderate to severe atopic dermatitis. Br J Dermatol 2009;161:1248-54. Crossref

32. Haeck I, van Velsen S, de Bruin-Weller M, Bruijnzeel-Koomen C. Bone mineral density in patients with atopic dermatitis. Chem Immunol Allergy 2012;96:96-9.  Crossref

33. Aalto-Korte K, Turpeinen M. Bone mineral density in patients with atopic dermatitis. Br J Dermatol 1997;136:172-5. Crossref

34. Pedreira CC, King E, Jones G, et al. Oral cyclosporin plus topical corticosteroid therapy diminishes bone mass in children with eczema. Pediatr Dermatol 2007;24:613-20. Crossref

35. van Velsen SG, Knol MJ, van Eijk RL, et al. Bone mineral density in children with moderate to severe atopic dermatitis. J Am Acad Dermatol 2010;63:824-31. Crossref

36. Andersen LF, Lande B, Trygg K, Hay G. Validation of a semi-quantitative food-frequency questionnaire used among 2-year-old Norwegian children. Public Health Nutr 2004;7:757-64. Crossref

37. Blum RE, Wei EK, Rockett HR, et al. Validation of a food frequency questionnaire in native American and Caucasian children 1 to 5 years of age. Matern Child Health J 1999;3:167-72. Crossref

38. Parrish LA, Marshall JA, Krebs NF, Rewers M, Norris JM. Validation of a food frequency questionnaire in preschool children. Epidemiology 2003;14:213-7. Crossref

39. Pearce MS, Salotti JA, Little MP, et al. Radiation exposure from CT scans in childhood and subsequent risk of leukaemia and brain tumours: a retrospective cohort study. Lancet 2012;380:499-505. Crossref

40. Mainz JG, Sauner D, Malich A, et al. Cross-sectional study on bone density-related sonographic parameters in children with asthma: correlation to therapy with inhaled corticosteroids and disease severity. J Bone Miner Metab 2008;26:485-92. Crossref

Child abuse is damaging to children’s physical health, emotional health, learning, and development.1 2 3 From time to time, there are media reports of severe child abuse that has required admission to an intensive care unit or resulted in death. A recent recommendation in March 2016 by a coroner following an inquest into the death of a 5-year-old child was the need for a careful risk assessment when handling cases of suspected child abuse.4

In Hong Kong, approximately 1000 children are admitted to hospitals each year for suspected child abuse. The abuse may be physical, sexual, or psychological; involve neglect; or consist of multiple abuses.5 Management of these children calls for multidisciplinary involvement. A Multi-Disciplinary Case Conference on Protection of Child with Suspected Abuse (MDCC) is recommended as stated in the Procedural Guide for Handling Child Abuse Cases of the Social Welfare Department of the Hong Kong Special Administrative Region (HKSAR) Government.6 Whether a case is abuse or not is concluded by the MDCC that involves doctors, nurses, psychologists, medical social workers, social workers from Social Welfare Department or non-governmental organisations, school personnel, and the police.

The Procedural Guide6 is under review. New procedures introduced in its recent revision have been implemented since December 2015. In Chapter 11 of the MDCC, a new standing conference agenda item on risk assessment was introduced and mandated. Managing professionals are advised to perform risk assessment on abuse. This risk assessment is vital when considering the nature of child abuse and the care of the child and family. Several assessment instruments or models to assess harm have been reviewed.7 8 Each has its own strengths and weaknesses. The Risk Assessment Matrix (developed by the California State University, Fresno, in 19909) has been quoted in the Procedural Guide for Handling Child Abuse Cases of the Social Welfare Department, HKSAR Government.6 The Risk Assessment Matrix has not been previously systematically used in MDCC in Hong Kong. Since 2015, the Social Welfare Department of HKSAR Government has recommended that systematic risk assessment be performed in MDCC for all cases. This study was performed to examine the relationship between risk factors in the Risk Assessment Matrix and MDCC outcome.

United Christian Hospital is a tertiary referral hospital that serves a paediatric population of around 110 000 in the Kwun Tong district in Hong Kong.10 Children with suspected child abuse are admitted to hospitals in Hong Kong for multidisciplinary management that includes work-ups by paediatrics, psychology, psychiatry, social work disciplines as well as community social work agencies, schools, and the police. An MDCC is held within 10 working days in which all involved disciplines participate to conclude the nature of abuse (case conference outcome) and the subsequent welfare plan for the child and family. This was a retrospective case series with internal comparison to investigate the risk factors and case conference outcome of children admitted with suspected abuse from January 2012 to December 2014. Ethics approval for the study was obtained from the Kowloon Central/Kowloon East Clusters Research Ethics Committee of the Hospital Authority.

All cases of suspected child abuse (coded per the ICD-9 system) within the study period were identified from discharge diagnosis using the Hospital Authority Clinical Data Analysis and Reporting System electronic database. The medical records, the MDCC investigation reports by various disciplines, and the MDCC meeting minutes were retrieved and retrospectively reviewed. Cases were categorised as ‘established’ (E), ‘high risk’ (HR), or ‘not established’ (NE) for child abuse, as determined in the MDCC. All E and HR cases were included in the positive case conference outcome group, and all NE cases were included in the comparison group. Cases with no MDCC were excluded from analysis.

In this study, the ratio of E+HR:NE cases was 198:67 (ie 3:1). Using this sample size, and assuming an odds ratio (OR) of >2 would be considered significant, the chance of detecting a significant difference at the 5% level was 65%.

Baseline demographic data, type of abuse, abusers, and relevant risk factors were collected for all cases. The Risk Assessment Matrix (developed by the California State University, Fresno, in 19909) adopted by the Social Welfare Department in their Procedural Guide for Handling Child Abuse Cases6 was used to associate risk factors with final categorisation. The full risk assessment form is shown in the Appendix. This assessment categorises risk factors for child abuse into 15 matrix domains to assess the child, parent/caretaker, and family situation. For each matrix domain, the level of risk is classified as ‘low’ (MLR), ‘intermediate’ (MIR), or ‘high’ (MHR). The matrix was discussed in detail among the authors before starting the study, and details of classification clarified. Classification was performed by one author only, thereby eliminating the possibility of inter-rater variability. Cases that were difficult to classify were discussed among authors and decisions were made by consensus. Association between risk categories in the matrix and final categorisation was reviewed by looking at the MIR + MHR category in relation to case conference outcomes. To further quantify the matrix, an empirical scoring system was devised, with 1 point for MLR, 2 points for MIR, and 3 points for MHR in each of the 15 matrix domains. For each assessed case, an aggregate score of 15 to 45 was possible.

Statistical analysis was performed using the Statistical Package for the Social Sciences (Windows version 23.0; IBM Corp, Armonk [NY], United States). Categorical data were compared using the Chi squared test or Fisher’s exact test (for cells <5), and OR with 95% confidence interval (CI) were calculated. Continuous variables were compared using the independent t test, Mann-Whitney U test, or one-way analysis of variance (for multiple groups). Multivariate logistic regression (stepwise strategy) was used to determine the effect of individual matrix domains on case conference outcome. The independent variables used in logistic regression analysis were the matrices that showed a significant association in the initial univariate analysis. To study the association between matrix scores and final categorisation, a receiver operating characteristic (ROC) curve was plotted with sensitivity and specificity calculations. A two-sided P value of ≤0.05 was considered significant.

It was hypothesised that (1) risk factors for child abuse are present in a higher proportion in the E/HR cases compared with the NE cases, and (2) the aggregate risk profile score is higher in the positive case conference outcome group than in the comparison group.

We identified 272 cases during the study period. After review of diagnosis and case notes, seven cases were excluded. For all excluded cases, no MDCC was held because they were judged to be inappropriate referrals for assessment of child abuse after initial careful assessment. Therefore, 265 cases were included in the study.

After multidisciplinary work-up, the case conference conclusion by MDCC showed that 46.0% (122/265) of cases were categorised as E, 28.7% (76/265) as HR, and 25.3% (67/265) as NE. There were ultimately 198 cases in the positive case conference outcome group (E+HR) and 67 in the comparison group (NE). Physical abuse cases accounted for 70.9% (188/265), and the percentages of sexual abuse, neglect, and multiple abuse (≥2 abuse categories) were 14.0%, 5.7%, and 9.4%, respectively. There were nine cases of psychological abuse (3 E and 6 HR), but they were also confirmed to be associated with other types of abuse (eg ‘physical + psychological’ or ‘neglect + psychological’). There were no cases of ‘isolated psychological abuse’ in this series (Table 1).

In most cases the abuser was identified as the mother (45.5%, 90/198), followed by the father (27.3%, 54/198), domestic helper (4.0%, 8/198), parent’s co-habitant (2.0%, 4/198), grandfather (1.5%, 3/198) or grandmother (1.5%, 3/198), internet friend (1.5%, 3/198), or stepfather (1.0%, 2/198) or stepmother (1.0%, 2/198). In 4.5% of cases, multiple abusers were identified, and in 5.1%, the abuser could not be identified. Other abusers accounted for 5.1% and included tutorial class teachers, mother’s friends, classmate or hostel peer, siblings, boyfriend, godmother, and other relatives.

Comparison of baseline demographic data showed no significant difference in gender, ethnicity, or mean age at presentation among the E, HR, and NE groups (Table 1). When the nature of abuse was compared, there was a higher percentage of physical abuse in the E and HR groups, but no significant difference in the percentage of psychological, multiple abuse, or neglect between groups. A significant difference was identified for sexual abuse, however, with the highest percentage in the NE group (10.7% vs 9.2% vs 25.4%; P=0.007). Several features were observed in this subgroup of sexual abuse as follows. Multidisciplinary investigations and physical examination were frequently not revealing. Children were often young and thus unable to speak with non-specific vulval or perineal redness or symptomatic vulvovaginitis. Child custody disputes, maternal emotional problems, or a child being cared for by multiple individuals were common features.

Univariate analysis for the MIR + MHR categories in each matrix domain showed significant correlation between MIR + MHR in the positive case conference outcome group (E + HR) for nine matrix domains, including Matrix 2, 3, 4, 5, 8, 9, 10, 11, and 14 (Table 2).

Logistic regression for these nine matrix domains showed significant correlation for Matrix 2, severity and/or frequency of abuse; Matrix 4, location of injuries; and Matrix 11, strength of family support systems (Table 3).

Using the devised scoring system of 1 point for MLR, 2 for MIR and 3 for MHR, an aggregate matrix score was calculated for each patient, with a minimum possible score of 15, and maximum possible score of 45. The aggregate matrix scores for both the positive case conference outcome group (E+HR) and comparison group (NE) followed a normal distribution (Fig 1). The mean aggregate matrix score was significantly different between the two groups with a higher mean score in the positive case conference outcome group (26.90 ± 3.57 vs 23.46 ± 2.98; P<0.005).

To estimate the association of the matrix scores with the risk of child abuse, an ROC curve was plotted using aggregate matrix score against E + HR cases (Fig 2). The area under the ROC curve was 0.78 (95% CI, 0.72-0.84), indicating good discrimination.

The sensitivity and specificity of different aggregate matrix scores are shown in Figure 2.

For this study, a matrix score that yielded a high sensitivity was preferred, in order to avoid missing cases of abuse. A cut-off aggregate matrix score of 23 would yield a sensitivity of 91.4% and specificity of 38.2% in relation to E + HR; a mean aggregate matrix score of 24 would yield a sensitivity of 84.8% and specificity of 48.5%.

Risk assessment is a critical process by which to assess the level of risk to a child suspected of being abused. Instruments used in risk assessment organise factors systematically to help describe the safety of such a child. These factors include characteristics of the reported abuse, the child, the caretakers, the family, and the environment of the child.7 8 11 12 13 Such assessment helps case analysis and decision making, and provides an important framework for case planning and subsequent service delivery.

Since December 2015, risk assessment in MDCC has been mandated in the Procedural Guide for Handling Child Abuse Cases of the Social Welfare Department, HKSAR Government.6 In this Procedural Guide, a risk assessment instrument (Risk Assessment Matrix developed by the California State University, Fresno, in 19909) was referred to and takes the form of a matrix that facilitates assessment by professionals of the level of risk for various abuse factors. This study examined the relationship between child abuse risk factors and MDCC outcome using this Risk Assessment Matrix.

There was no significant difference in demographic data among the three groups (E, HR, and NE; Table 1). A statistical difference in the presence of child sexual abuse was found between the positive case conference outcome group (E+HR) and the comparison group (NE), with a higher proportion of children in the comparison group affected. Future study to analyse characteristic features of the NE group would aid understanding of sexual abuse cases that present to hospitals in Hong Kong.

There was no ‘isolated’ psychological abuse in this series. All psychological abuses occurred with multiple abuses. Psychological abuse is easily missed as there is often no physical sign to arouse suspicion. All cases in this series came to light during the work-up for other forms of abuse. In 2015, there were only seven cases of psychological abuse among the 874 newly reported child abuse cases in Hong Kong.14 Psychological abuse is underdiagnosed in our locality and this calls for sensitivity among professionals when handling abuse cases.

On characteristics of abusers, parents, especially mothers, were the most prevalent abusers. This finding is consistent with previous studies.12 15 16 17 Certain parental characteristics have been identified as important risk factors for child abuse, for example, parental low mood, marital conflict precipitating emotional problems, parental low education or economic status, poor social support, and parenting stress due to handling a child’s disruptive behaviour.12 15 16 17 18

Logistic regression analysis revealed three factors that were significant for established or high risk of child abuse (E or HR): (1) Matrix 2: severity and/or frequency of suspected physical or sexual abuse, (2) Matrix 4: location of injuries, and (3) Matrix 11: strength of family support systems (Table 3).

History of child abuse, and severity and frequency of abuse are known risk factors for recurrence of abuse. Child abuse victims may not experience abuse as a one-off event. Further, there was evidence of escalation in abuse severity in recurrent abuse victims.19 20 Corporal punishment is commonly adopted by Chinese parents as a method of child discipline, and severe physical punishment warranting medical attention or hospital admission has been reported in 3% to 9% of children.15 18 Only 1% of abuse cases are reported and managed.15 Contributing factors for underreporting include cultural acceptance of corporal punishment, low public awareness, and lack of victim support during the disclosure process.

Head and neck injury was regarded as severe physical injury compared with injury to limbs and corporal body parts.18 A review of literature revealed that abusive bruises are found predominantly on the head and neck, especially on the ear, neck, and cheeks—all sites that are unlikely to be affected by accidental injury. Areas such as the forearms, upper limbs, and adjoining area of the trunk, or outside thigh may indicate ‘defensive bruising’ when the child tries to avoid being hit.21 Head and neck injuries such as abusive head injury, contusions of the head or neck, are well known to cause deleterious effects, even mortality.22

Families with poor social support, social isolation, and geographical isolation are known to be at increased risk and severity of child abuse.16 17 19 22 Social isolation was more common among single parents or immigrants.15 Both a low level of real and perceived social support has been shown to be potential risks for child maltreatment.15 16 17 On the contrary, social support is a protective factor for child abuse.23 Perceived social support has been reported to moderate parents’ own experience of abuse and the potential risk of abuse of their own children.16 Parental support can be offered by child care or foster care services, targeted support programmes for families at risk or young families with a newborn, parental counselling service, and extra support to vulnerable children with special needs.23

Six other matrix domains were significantly related with case conference outcome in univariate analysis but not in logistic regression analysis (Tables 2 and 3). They were Matrix 3, 5, 8, 9, 10 and 14. Another six risk factors were not statistically related to case conference outcome; these included Matrix 1, 6, 7, 12, 13, and 15. All risk factors in these domains have been shown in previous studies to be related to child abuse.11 12 13 15 16 17 Possible explanations for the absence of a significant relationship between risk factors in these 12 domains and case conference outcome in logistic regression analysis include an aggregate effect of risk factors that may not be significant on their own but factor co-occurrence is contributory. Other possible explanations include presence of mitigating factors such as a protective relative, a child already in supportive placement, the presence of legal enforcement or a child under a care order, or because of a small subgroup number within individual risk factors.

For the aggregate effect of risk factors, an ROC curve was plotted using aggregate matrix score against case conference outcome (Figs 1 and 2). As the Risk Assessment Matrix is used as a risk assessment tool for child abuse, it is vital that it detects most abuse cases. We chose a score that yields a high sensitivity and high positive predictive value whilst accepting a lower specificity. Using a score of 23 (sensitivity 91.4%, positive predictive value 0.85, specificity 38.2%) or 24 (sensitivity 84.8%, positive predictive value 0.8, specificity 48.5%) ensured that most child abuse cases were identified. The high sensitivity indicates that most cases of E and HR child abuse would be correctly identified in MDCC. A welfare plan could then be formulated to protect the child and help the family to prevent further abuse. The low specificity, however, meant that a relatively large number of ‘non–child abuse’ cases could be subject to unnecessary investigations, leading to an increased workload for all parties involved and stress to the family. Nonetheless, a highly sensitive cut-off is important to avoid a false-negative result and missing a genuine case of child abuse that may have serious or even fatal consequences.

In a recent death inquest, the importance of risk assessment was strongly emphasised by the coroner.4 The aggregate matrix score offers a reference to alert professionals in handling suspected child abuse cases. A matrix score of >23 calls for increased vigilance and careful planning, especially in situations such as making a decision about hospital discharge before MDCC. Further, because job placements of disciplines such as social work or legal enforcement are often rotation-based rather than long-term specialist-focused, where experience and professional judgement are important cumulative assets, a systematic risk assessment using objective scores serves as a practical tool and as a warning mechanism in abuse handling, especially for the less-experienced professionals.

This study has some limitations. In Hong Kong, reported cases of child abuse are only the tip of the iceberg.15 Subjects in this study were hospitalised children in a regional hospital setting, and results of this retrospective study cannot be generalised to the territory. The Fresno model has previously been considered a model with low validity and inter-rater reliability.7 As with other consensus-based risk assessment instruments, the rating of risks in the matrix domains will invariably involve a degree of subjectivity.7 8 This was minimised in this study by our further defining situations with objective measures. For example, for domain 10, intermediate risk was defined as a reported case but subsequently concluded as not an established child abuse case to be followed up by a school social worker or Integrated Family Services Centre. High risk was defined as a history of established child abuse in the past. For domain 14, insufficient income was defined as receipt of Comprehensive Social Security Assistance. Recent change in marital or relationship status was defined as parents in divorce proceedings, child in a custody dispute, or active marital discord causing emotional outbursts. It is hoped that with training and further refining of the matrix contents to fit the local culture, the inter-rater reliability and reproducibility of the Fresno tool can be improved. Nevertheless other risk assessment instruments can also be examined for local use.

The social structure and culture of a society keeps changing. Up-to-date studies are required to examine child abuse risk profiles. A prospective multicentre study is valuable for development of a local risk assessment tool. With the implementation of changes in the Procedural Guide for Handling Child Abuse Cases,6 a systematic risk assessment will facilitate investigative procedures and improve safeguarding of vulnerable children.

Three matrix risk factors in the Risk Assessment Matrix were significantly associated with child abuse—severity and/or frequency of suspected physical or sexual abuse (Matrix 2), location of injuries (Matrix 4), and strength of family support systems (Matrix 11). Further, other risk factors in the matrix, although not significant in logistic regression analysis, showed good association with child abuse case conference outcomes in univariate analysis. A risk assessment framework facilitates case analysis, and guides decision making and case planning such that appropriate service delivery is ensured. Using the devised scoring system of the referenced Risk Assessment Matrix, an aggregate matrix score of ≥23 should arouse suspicion of all professionals when managing child abuse.

All authors have disclosed no conflicts of interest.

1. Norman RE, Byambaa M, De R, Butchart A, Scott J, Vos T. The long-term health consequences of child physical abuse, emotional abuse, and neglect: a systematic review and meta-analysis. PLoS Med 2012;9:e1001349. Crossref

2. Gilbert R, Widom CS, Browne K, Fergusson D, Webb E, Janson S. Burden and consequences of child maltreatment in high-income countries. Lancet 2009;373:68-81. Crossref

3. Felitti VJ, Anda RF, Nordenberg D, et al. Relationship of childhood abuse and household dysfunction to many of the leading causes of death in adults. The Adverse Childhood Experiences (ACE) study. Am J Prev Med 1998;14:245-58. Crossref

4. Siu J. Hong Kong government urged to amend guide on handling child abuse in coroner’s case involving death of boy who probably ingested Ice. South China Morning Post 2016 Mar 16.

5. Clinical Data Analysis and Reporting System, Hong Kong Hospital Authority. Accessed 16 Nov 2016.

6. Social Welfare Department of the Hong Kong SAR Government. Procedural Guide for Handling Child Abuse Cases 2015. Available from: http://www.swd.gov.hk/en/index/site_pubsvc/page_family/sub_fcwprocedure/id_1447/. Accessed 16 Nov 2016.

7. D’Andrade A, Austin MJ, Benton A. Risk and safety assessment in child welfare: instrument comparisons. J Evid Based Soc Work 2008;5:31-56. Crossref

8. Barlow J, Fisher JD, Jones D. Systematic review of models of analysing significant harm. Research report DFE-RR199. London: Department for Education; 2012.

9. California risk assessment curriculum for child welfare services, CSU Fresno, Child Welfare Training Project. Sponsored and funded by the California State Department of Social Service; 1990.

10. Census and Statistics Department. Population and household statistics analysed by District Council district. Hong Kong, Hong Kong SAR Government; 2016.

11. Milner JS. Assessing physical child abuse risk: the child abuse potential inventory. Clin Psychol Rev 1994;14:547-83. Crossref

12. Begle AM, Dumas JE, Hanson RF. Predicting child abuse potential: an empirical investigation of two theoretical frameworks. J Clin Child Adolesc Psychol 2010;39:208-19. Crossref

13. Chan YC, Lam GL, Chun PK, So MT. Confirmatory factor analysis of the Child Abuse Potential Inventory: results based on a sample of Chinese mothers in Hong Kong. Child Abuse Negl 2006;30:1005-16. Crossref

14. Social Welfare Department. Child Protection Registry statistical report 2015. Hong Kong: Hong Kong SAR Government; 2015.

15. Study on child abuse and spouse battering. Report on findings of household survey. Hong Kong SAR: Department of Social Work and Social Administration, The University of Hong Kong; 2005.

16. Yoon AS. The role of social support in relation to parenting stress and risk of child maltreatment among Asian American immigrant parents [dissertation]. US: University of Pennsylvania; 2013.

17. Brown J, Cohen P, Johnson JG, Salzinger S. A longitudinal analysis of risk factors for child maltreatment: findings of a 17-year prospective study of officially recorded and self-reported child abuse and neglect. Child Abuse Negl 1998;22:1065-78. Crossref

18. Leung PW, Wong WC, Chen WQ, Tang CS. Prevalence and determinants of child maltreatment among high school students in Southern China: a large scale school based survey. Child Adolesc Psychiatry Ment Health 2008;2:27. Crossref

19. Thackeray J, Minneci PC, Cooper JN, Groner JI, Deans KJ. Predictors of increasing injury severity across suspected recurrent episodes of non-accidental trauma: a retrospective cohort study. BMC Pediatr 2016;16:8. Crossref

20. Deans KJ, Thackeray J, Groner JI, Cooper JN, Minneci PC. Risk factors for recurrent injuries in victims of suspected non-accidental trauma: a retrospective cohort study. BMC Pediatr 2014;14:217. Crossref

21. Maguire S. Which injuries may indicate child abuse? Arch Dis Child Educ Pract Ed 2010;95:170-7. Crossref

22. Kemp AM. Abusive head trauma: recognition and the essential investigation. Arch Dis Child Educ Pract Ed 2011;96:202-8. Crossref

23. Chan KL. Study on child-friendly families: Immunity from domestic violence. Hong Kong SAR: Department of Social Work and Social Administration, The University of Hong Kong; 2008.

Over the past couple of decades, there has been a move towards fast-track surgery designed to reduce postoperative morbidities and length of hospital stay.1 Laparoscopic methods for colonic surgery have accelerated postoperative recovery by reducing the time required for bowel function recovery and enhancing postoperative mobilisation.2 Postoperative ileus, however, remains a common reason for prolonged hospital stay following major abdominal surgery. Although its pathophysiology is multifactorial, use of opioids as postoperative analgesia is thought to contribute to the problem.3 4 Therefore, safe and effective postoperative pain control with minimal use of opioids is essential to enhance recovery.5

The advantages of continuous infusion of thoracic epidural analgesia (TEA) compared with intravenous (IV) patient-controlled analgesia with opioid have been studied. The results show that TEA significantly improves early analgesia requirement following laparoscopic colectomy with an opioid-sparing effect. Nonetheless TEA is associated with other adverse reactions such as urinary retention, hypotension, epidural haematoma, and abscess formation.6 Intravenous infusion of lignocaine has emerged in recent years as a feasible, cost-effective, and safe method to provide postoperative analgesia.7 Recent randomised controlled trials have shown that the combined analgesic, anti-inflammatory, and antihyperalgesic properties of IV lignocaine improve outcomes and shorten hospital stay following colorectal surgery.8 There is level I (PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-Analyses) evidence that IV lignocaine infusions are opioid-sparing and significantly reduce pain scores at rest and during activity, nausea, vomiting, duration of ileus after abdominal surgery, and length of hospital stay. Peri-operative IV administration of lignocaine also has a preventive analgesic effect following a wide range of operations.9 10 11

Currently there is no literature about this perioperative pain control modality in the Chinese patients. In the following account, we present a case series of Chinese patients who underwent laparoscopic colorectal surgery and received a peri-operative IV lignocaine infusion.

We reviewed cases of patients who underwent elective laparoscopic resection of colorectal cancer and received a lignocaine infusion as postoperative analgesia between September 2012 and May 2015 at North District Hospital in Hong Kong. This study aimed to determine whether postoperative IV lignocaine infusion would provide adequate analgesia, shorten the duration of postoperative ileus, reduce postoperative complications, enhance rehabilitation, and shorten hospital stay. This study was done in accordance with the principles outlined in the Declaration of Helsinki.

All patients were assessed preoperatively by an anaesthetist to exclude any contra-indications to use of IV lignocaine. Routine consent for anaesthesia was obtained with clear choices offered for postoperative analgesia and the relevant risks explained to the patient. The choices for postoperative analgesia included epidural analgesia, IV lignocaine infusion, and IV patient-controlled analgesia with morphine. Intravenous lignocaine was offered when patients refused or were contra-indicated for epidural analgesia. If patients were not suitable for either epidural analgesia or IV lignocaine, IV patient-controlled analgesia with morphine was offered. The anaesthetic technique was standardised for all patients.

All patients received an IV bolus injection of lignocaine 1.5 mg/kg over 20 minutes on induction followed by a continuous infusion of 1.5 mg/kg/h intra-operatively. The 1% lignocaine infusion was continued at a rate of 1 mg/kg/h for 24 hours postoperatively, delivered through a GemStar infusion device with the fixed calculated dose set up by the case anaesthetist. For safety reasons, the lignocaine infusion was connected to a dedicated IV line to avoid accidental bolus administration. General anaesthesia was induced with fentanyl 1-2 µg/kg, propofol 2-3 mg/kg, and cisatracurium 0.15-0.2 mg/kg for intubation. Anaesthesia was maintained with oxygen in room air or nitrous oxide and isoflurane or sevoflurane at an end-tidal anaesthetic concentration of approximately 1 minimal alveolar concentration. Ketorolac 15-30 mg was administered on induction if not contra-indicated clinically. Intravenous tramadol 50-100 mg and morphine was used intra-operatively for analgesia as decided by the list anaesthetist. Wound infiltration of local anaesthetic, 0.25% levobupivacaine 20 mL, was administered by the surgeon at the end of surgery.

Patients who had colorectal cancer and underwent elective laparoscopic colorectal resection were recruited into the study. All patients had colorectal cancer but the surgical procedure performed depended on the location of the tumour and the international standard. The surgeries included: laparoscopic right hemicolectomy (n=2), laparoscopic left hemicolectomy (n=3), laparoscopic sigmoidectomy (n=5), laparoscopic anterior resection of rectum (n=1), laparoscopic lower anterior resection with total mesorectal excision and stoma formation (n=4), and laparoscopic abdominoperineal resection (n=1). All patients had four to five small incisions for the laparoscopic procedure together with one larger 6- to 8-cm abdominal incision for specimen retrieval. For the patient with laparoscopic abdominoperineal resection, a larger wound for specimen retrieval was made over the perineal region instead of the abdomen.

All patients were prescribed regular oral paracetamol 500 mg to 1 g 3 to 4 times per day. Regular oral diclofenac SR 100 mg daily for 3 days was prescribed if not contra-indicated. As required, IV tramadol 50 mg every 6 to 8 hours was given if pain was not adequately controlled. Rescue subcutaneous morphine was prescribed in the protocol for severe uncontrolled pain.

All postoperative data were collected prospectively. The acute pain service and ward nurses followed the clinical plan that was devised by both the surgical and pain team.

After surgery, a categorical pain score (divided into none, mild, moderate, or severe pain) was obtained immediately in the postoperative care unit by recovery nurses. After the patient was discharged to the ward, pain scores were obtained by ward nurses on a numerical rating scale at rest hourly for 24 hours until lignocaine infusion was stopped. Patients would be reviewed by acute pain management team before lignocaine infusion was stopped and pain scores on postoperative day 1 were obtained at rest and during mobilisation. The numerical rating scale scored pain from 0 to 10 with 0 being no pain and 10 being the worst pain imaginable. Pain scores are continuous variables and are presented as median (interquartile range [IQR]) scores against time. Patient satisfaction score from 0 to 10 was also assessed by the acute pain management team. The presence of nausea, vomiting, dizziness, and other possible side-effects was documented. Intra-operative and postoperative analgesic consumption was recorded. All patients were monitored by cardiac monitor intra-operatively by anaesthetists and postoperatively in the recovery room by nurses. When patients were discharged to the ward, they were monitored for the next 24 hours until the end of IV lignocaine infusion with vital signs recorded every hour, including blood pressure, pulse, saturation, and continuous cardiac monitoring. There was no recorded cardiac arrhythmia event noted for any patient.

Return of bowel function was assessed by calculating the time from end of surgery to the passage of first flatus and first bowel opening. Postoperative rehabilitation was assessed by the time taken to tolerate diet and achieve full mobilisation and the length of hospital stay. These data are expressed as median (IQR) scores.

Sixteen patients were studied with a mean (± standard deviation) age of 66 ± 10 years. All were classified as American Society of Anesthesiologists grade I to III. Demographic data and duration of surgery are shown in Table 1.

During IV lignocaine infusion, four patients experienced nausea, one vomited, and two complained of mild dizziness. No serious adverse reactions were reported. All patients tolerated and completed the infusion of lignocaine.

In the postoperative care unit, most patients experienced none or mild pain. Only one patient complained of severe pain and required a fentanyl bolus for rescue analgesia. The self-reported pain scores are shown in Table 2. In addition to regular paracetamol, five patients requested IV tramadol for rescue analgesia in the first 24 hours postoperatively; these patients received tramadol 50-150 mg. No patient requested morphine during the first 24 hours postoperatively. Of the 16 patients, 11 showed overall satisfaction with the analgesia with median satisfaction score of 7.5 (7-9).

As seen in Table 3, the median times to first flatus and first bowel opening in the postoperative period were 21 (18-35) hours and 3 (1-3) days, respectively. The median times to tolerating diet and mobilisation were 1 (1-1) day and 2 (2-3) days, respectively. No patient had postoperative ileus. Only one patient had acute retention of urine that delayed discharge from hospital. Three other patients had a prolonged hospital stay due to social problems.

There was no documented postoperative arrhythmia for any patient.

Although this is a small case-series review, we have shown that lignocaine infusion is a safe and feasible means of postoperative pain control for patients undergoing laparoscopic colorectal resection. There was no major or serious adverse reaction such as cardiac arrhythmia during the lignocaine infusion. We also demonstrated that lignocaine infusion provided effective analgesia over the first 24 hours with acceptable pain score, low rescue opioid consumption, and good patient satisfaction score. Our results are consistent with the literature. Harvey et al12 observed that pain scores were decreased when a lignocaine infusion was administered compared with a group who received IV infusion of normal saline. Kaba et al13 also demonstrated that their lignocaine group required 50% less opioid during the first 24 hours postoperatively. Similar results were reported in other randomised controlled trials demonstrating that IV lignocaine has an opioid-sparing effect as an adjuvant analgesic.7 14 A recent meta-analysis by McCarthy et al15 examined the overall efficacy of IV lignocaine on postoperative analgesia and recovery from surgery in patients undergoing various surgical procedures. It concluded that IV lignocaine infusion in the perioperative period has clear advantages in patients undergoing abdominal surgery in terms of both pain control and bowel motility.

Our study also observed that IV lignocaine resulted in rapid recovery of bowel function and mobilisation. The median time for return of flatus and ability to tolerate an oral diet was within 24 hours. The median (IQR) time for bowel opening was 3 (1-3) days. These results are similar to the findings of Kaba et al13 who showed that lignocaine infusion improved postoperative bowel function. In that study, defaecation occurred almost 1 day earlier in the lignocaine group compared with the group who received normal saline. The reasons for postoperative ileus are multifactorial, including use of opioid analgesia, the sympathetic response, and visceral inflammatory response resulting from surgery.16 A lignocaine infusion may shorten the time to bowel opening by decreasing opioid use, limiting the inflammatory response, and having a direct inhibitory effect on the sympathetic nervous system of the mesenteric nervous plexus resulting in enhanced bowel contractility.17

A meta-analysis showed that continuous IV administration of lignocaine significantly reduces the length of hospital stay when compared with controls.17 In our study, however, the median hospital stay was 6 days, similar to our usual experience. We are evaluating the possible reasons for the lack of impact on hospital stay. One of the reasons may be related to patient expectations and preference for a longer hospital stay after major surgery. Another possible reason is the similar rehabilitation care pathway for the two groups of patients that when strictly followed tended to negate the advantages of IV lignocaine.

This review shows promising results demonstrating that IV lignocaine is a safe and effective postoperative analgesia in a Chinese population. It also provides comparable outcomes to those reported worldwide that postoperative lignocaine can provide a beneficial rehabilitation effect for patients who have undergone laparoscopic colorectal surgery. This provides a good platform from which to design a randomised controlled trial in the Chinese population.

The authors declared no conflicts of interest in this study.

1. Kehlet H, Dahl JB. Anaesthesia, surgery, and challenges in postoperative recovery. Lancet 2003;362:1921-8. Crossref

2. Reza MM, Blasco JA, Andradas E, Cantero R, Mayol J. Systematic review of laparoscopic versus open surgery for colorectal cancer. Br J Surg 2006;93:921-8. Crossref

3. Baig MK, Wexner SD. Postoperative ileus: a review. Dis Colon Rectum 2004;47:516-26. Crossref

4. Taguchi A, Sharma N, Saleem RM, et al. Selective postoperative inhibition of gastrointestinal opioid receptors. N Engl J Med 2001;345:935-40. Crossref

5. Kuhry E, Schwenk W, Gaupset R, Romild U, Bonjer J. Long-term outcome of laparoscopic surgery for colorectal cancer: a cochrane systematic review of randomized controlled trials. Cancer Treat Rev 2008;34:498-504. Crossref

6. Senagore AJ, Delaney CP, Mekhail N, Dugan A, Fazio VW. Randomized clinical trial comparing epidural anaesthesia and patient-controlled analgesia after laparoscopic segmental colectomy. Br J Surg 2003;90:1195-9. Crossref

7. Koppert W, Weigand M, Neumann F, et al. Perioperative intravenous lidocaine has preventive effects on postoperative pain and morphine consumption after major abdominal surgery. Anesth Analg 2004;98:1050-5. Crossref

8. Herroeder S, Pecher S, Schönherr ME, et al. Systemic lidocaine shortens length of hospital stay after colorectal surgery: a double-blinded, randomized, placebo-controlled trial. Ann Surg 2007;246:192-200. Crossref

9. Vigneault L, Turgeon AF, Côté D, et al. Perioperative intravenous lignocaine infusion for postoperative pain control: a meta-analysis of randomized control trials. Can J Anaesth 2011;58:22-37. Crossref

10. Sun Y, Li T, Wang N, Yun Y, Gan TJ. Perioperative systemic lidocaine for postoperative analgesia and recovery after abdominal surgery: a meta-analysis of randomized control trials. Dis Colon Rectum 2012;55:1183-94. Crossref

11. Barreveld A, Witte J, Chahal H, Durieux ME, Strichartz G. Preventive analgesia by local anesthetics: the reduction of postoperative pain by peripheral nerve blocks and intravenous drugs. Anesth Analg 2013;116:1141-61. Crossref

12. Harvey KP, Adair JD, Isho M, Robinson R. Can intravenous lidocaine decrease postsurgical ileus and shorten hospital stay in elective bowel surgery? A pilot study and literature review. Am J Surg 2009;198:231-6. Crossref

13. Kaba A, Laurent SR, Detroz BJ, et al. Intravenous lidocaine infusion facilitates acute rehabilitation after laparoscopic colectomy. Anesthesiology 2007;106:11-8. Crossref

14. Groudine SB, Fisher HA, Kaufman RP Jr, et al. Intravenous lidocaine speeds the return of bowel function, decreases postoperative pain, and shortens hospital stay in patients undergoing radical retropubic prostatectomy. Anesth Analg 1998;86:235-9. Crossref

15. McCarthy GC, Megalla SA, Habib AS. Impact of intravenous lidocaine infusion on postoperative analgesia and recovery from surgery: a systematic review of randomized controlled trials. Drugs 2010;70:1149-63. Crossref

16. Luckey A, Livingston E, Taché Y. Mechanisms and treatment of postoperative ileus. Arch Surg 2003;138:206-14. Crossref

17. Marret E, Rolin M, Beaussier M, Bonnet F. Meta-analysis of intravenous lidocaine and postoperative recovery after abdominal surgery. Br J Surg 2008;95:1331-8. Crossref